University of Arkansas, Fayetteville

ScholarWorks@UARK

Graduate Theses and Dissertations

5-2019

The Effect of Using Technology on Students’ Understanding in

Calculus and College Algebra
Razieh Shahriari
University of Arkansas, Fayetteville

Follow this and additional works at: https://scholarworks.uark.edu/etd

Part of the Educational Assessment, Evaluation, and Research Commons, Higher Education
Commons, and the Science and Mathematics Education Commons

Citation
Shahriari, R. (2019). The Effect of Using Technology on Students’ Understanding in Calculus and College
Algebra. Graduate Theses and Dissertations Retrieved from https://scholarworks.uark.edu/etd/3264

This Dissertation is brought to you for free and open access by ScholarWorks@UARK. It has been accepted for
inclusion in Graduate Theses and Dissertations by an authorized administrator of ScholarWorks@UARK. For more
information, please contact scholar@uark.edu.
 The Effect of Using Technology on Students’ Understanding in
Calculus and College Algebra

A dissertation submitted in partial fulfillment

of the requirements for the degree of
Doctor of Philosophy in Mathematics

by

Razieh Shahriari
Shahid Rajaei University
Bachelor of Science in Mathematics, 2010
University of Arkansas
Master of Science in Mathematics, 2015

May 2019
University of Arkansas

This thesis is approved for recommendation to the Graduate Council.

Bernard L. Madison, Ph.D. Shannon W. Dingman, Ph.D.

Committee Member Committee Member

Mark Arnold, Ph.D. Nama Namakshi, Ph.D.

Committee Member Committee Member
Abstract
This mixed qualitative and quantitative methods study addressed the effect of technology

on college algebra and survey of calculus students’ understanding. This research study was

conducted in fall 2016 on eight college algebra classes with a total of 315 students, and in summer

2017, on two survey of calculus classes with a total of 40 students at the University of Arkansas.

Several sources were used to collect data. A pre- and post- student attitude survey was

administered during the first and last week of the semester for both college algebra and survey of

calculus courses. Students’ scores and paper work on three written tests (review test 1, review test

2 and concept test) in college algebra and students’ scores and paper work on two written tests

(review test 1 and review test 2) in survey of calculus were collected. The concept test was the

only paper test normally administered in college algebra. Quantitative and qualitative data analysis

enabled discussion of the effect of technology on students’ understanding and organization of their

work. This research study was guided by the following research questions.

1. How does the use of technology affect college algebra and calculus students’ understanding

and performance?

2. What areas of college algebra and calculus are affected more by technology?

3. How does using technology affect the organization of college algebra and calculus students’

written work?

4. Does the use of technology positively impact college algebra and calculus students’

attitudes toward their mathematics skills?

The results from the study exposed evidence that use of technology (handheld graphing

calculators, online graphing utility Desmos, and smartphone apps) in teaching and learning
increased college algebra students’ understanding of several concepts such as domain, vertical and

horizontal asymptotes, end behavior of a function, and logarithmic functions. In addition, college

algebra students’ skills such as logical reasoning, use of graph, organization including written

order, and correct use of notation and symbols were significantly increased when they used

technology. Survey of calculus students’ understanding increased in several topics such as finding

maximum/minimum for two variable functions, limits, and definite integrals when they used

technology in their class activities and on written tests.

Acknowledgements

I would like to thank people who supported me by any means through this accomplishment.

Foremost I would like to express my special thanks to my advisor, Dr. Bernard Madison, for all

his help, guidance, enthusiasm, motivation and immense patronage. He trusted me and has given

me the freedom to pursue my research dependently and helped me not deviate from the core of my

research.

I would like to thank other committee members, Dr. Mark Arnold, Dr. Shannon Dingman,

and Dr. Nama Namakshi for their precious advices and insights in my research studies. Especially

Dr. Namakshi who provided her invaluable vision on the logarithm project.

I have great pleasure to acknowledge my gratitude to the coordinator and instructors of

college algebra and survey of calculus courses, Ms. Kruse, Dr. Erdem, Mr. Mutanguha and Mr.

Keyton who agreed that their classes be subjects of this study and for their time for giving surveys

to the students and patiently helping me in collecting data.

In addition, I would like to thank Dr.Tipton, Dr. Zmhang, and Dr. Datta for discussion and

introducing me to some of the statistical methods I used in this dissertation.

I would like to thank all the students who participated in this study and filled out the surveys

and attended the interviews.

I would like to thank University of Arkansas for giving the permission to perform this

research study.

Finally, my deep and sincere gratitude goes to my beloved husband Dr. Mehrshad

Mehboudi, my son Ali, my family and friends for their continuous support, love, and help during
these years. I am forever indebted to family who supported me emotionally and encouraged me to

seek higher goals.

‫‪In loving memory of my sister Mehri Shahriari. Her memory is always with me and I live‬‬

‫‪through her eyes.‬‬

‫ای ﻣﮭرﺑﺎن ﺗر از ﺑرگ در ﺑوﺳﮫ ھﺎی ﺑﺎران‬

‫ﺑﯾداری ﺳﺗﺎره در ﭼﺷم ﺟوﯾﺑﺎران‬
‫آﯾﯾﻧﮫ ﻧﮕﺎھت ﭘﯾوﻧد ﺻﺑﺢ و ﺳﺎﺣل‬
‫ﻟﺑﺧﻧد ﮔﺎھﮕﺎھت ﺻﺑﺢ ﺳﺗﺎره ﺑﺎران‬
‫ﺑﺎز آ ﮐﮫ در ھواﯾت ﺧﺎﻣوﺷﯽ ﺟﻧوﻧم‬
‫ﻓرﯾﺎدھﺎ ﺑراﻧﮕﯾﺧت از ﺳﻧﮓ ﮐوھﺳﺎران‬
‫ای ﺟوﯾﺑﺎر ﺟﺎری !زﯾن ﺳﺎﯾﮫ ﺑرگ ﻣﮕرﯾز‬
‫ﮐﺎﯾﻧﮕوﻧﮫ ﻓرﺻت از ﮐف دادﻧد ﺑﯾﺷﻣﺎران‬
‫ﮔﻔﺗﯽ ﺑﮫ روزﮔﺎری ﻣﮭری ﻧﺷﺳﺗﮫ ﮔﻔﺗم‬
‫ﺑﯾرون ﻧﻣﯽ ﺗوان ﮐرد ﺣﺗﯽ ﺑﮫ روزﮔﺎران‬
‫ﺑﯾﮕﺎﻧﮕﯽ ز ﺣد رﻓت ای آﺷﻧﺎ ﻣﭘرھﯾز‬
‫زﯾن ﻋﺎﺷﻖ ﭘﺷﯾﻣﺎن ﺳرﺧﯾل ﺷرﻣﺳﺎران‬
‫ﭘﯾش از ﻣن و ﺗو ﺑﺳﯾﺎر ﺑودﻧد و ﻧﻘش ﺑﺳﺗﻧد‬
‫دﯾوار زﻧدﮔﯽ را زﯾﻧﮕوﻧﮫ ﯾﺎدﮔﺎران‬
‫اﯾن ﻧﻐﻣﮫ ﻣﺣﺑت‪ ،‬ﺑﻌد از ﻣن و ﺗو ﻣﺎﻧد‬
‫ﺗﺎ در زﻣﺎﻧﮫ ﺑﺎﻗﯾﺳت آواز ﺑﺎد و ﺑﺎران‬
‫دﮐﺗر ﻣﺣﻣدرﺿﺎ ﺷﻔﯾﻌﯽ ﮐدﮐﻧﯽ‬
Table of Contents

Overview ......................................................................................................................................... 1
Chapter 1: Introduction ................................................................................................................... 3
1.1 Purpose of the study ......................................................................................................... 4

1.2 Statement of the problem ................................................................................................. 4

1.3 Research questions ........................................................................................................... 5

1.4 Theoretical framework ..................................................................................................... 5

1.5 Significance of the research study.................................................................................... 7

1.6 Summary .......................................................................................................................... 8

Chapter 2: Literature review ........................................................................................................... 9

2.1 Technology in education .................................................................................................. 9

2.2 Technology and theories of learning .............................................................................. 10

2.3 Overview of different types of graphing calculators ...................................................... 11

2.4 Research on the effectiveness of using graphing calculators ......................................... 12

2.5 The effect of graphing calculators on students’ achievement ........................................ 12

2.6 The effect of graphing calculators on students’ learning skills ...................................... 13

2.7 Effect of graphing calculators on students’ performance............................................... 15

2.8 Students’ written work in the presence of technology ................................................... 16

2.9 Tablet and smartphone apps in education ...................................................................... 17

2.10 Mathematical apps in secondary and college courses .................................................... 19

 2.11 Summary..................................................................................................................... 21

Chapter 3 Methodology ................................................................................................................ 23

3.1 Data collection instruments ............................................................................................ 23

3.2 Participants ..................................................................................................................... 24

3.3 Student attitude survey ................................................................................................... 25

3.4 Supplementary information on college algebra course content ..................................... 26

3.5 Concept test, review tests 1 and 2 .................................................................................. 27

3.6 Teaching college algebra with the aid of technology..................................................... 29

3.7 Supplementary information of survey on calculus ......................................................... 35

3.8 Teaching calculus with the aid of a graphing calculator and Desmos ........................... 36

3.9 Primary data collection................................................................................................... 39

3.10 Primary evaluations ........................................................................................................ 42

3.11 Summary......................................................................................................................... 44

Chapter 4 Data analysis and results .............................................................................................. 45

4.1 Quantitative data analysis and result for college algebra ............................................... 46

4.2 Quantitative results of the concept tests ......................................................................... 49

4.3 Qualitative data analysis and result for college algebra ................................................. 52

4.4 Proportional stratified random sampling ........................................................................ 54

4.5 Validating the analytic rubric .......................................................................................... 57

4.6 Guideline and memo for using the analytic rubric over RT1 ........................................... 57
4.7 Memos ............................................................................................................................ 59

4.8 The result of the analytic rubric applied to RT1 and RT2 ............................................. 60

4.9 The result of t-test over mean score obtained using the analytic rubric......................... 61

4.10 Estimating the population mean and confidence interval of the analytic rubric ............ 63

4.11 A detailed qualitative description and discussion of students’ review tests ................... 66

4.12 Comparing Sara’s work for Q4 in RT1 and Q1 in RT2 .................................................. 72

4.13 The result of the analytic rubric applied to the concept test ............................................ 73

4.14 The result of the analytic rubric applied to treatment group .......................................... 78

4.15 Estimating the population mean and confidence interval ............................................... 79

4.16 A detailed qualitative description of students’ concept tests .......................................... 81

4.17 The result of student attitude survey (SAS) .................................................................... 85

4.18 Summary of results of pre- and post-survey in college algebra...................................... 95

4.19 Results of pre- and post-survey in survey of calculus .................................................... 97

4.20 Quantitative data analysis and result in survey of calculus ............................................. 97

4.21 The result of the analytic rubric applied to RT1 and RT2 ............................................ 100

4.22 Discussion of RT1 and RT2 for survey of calculus ...................................................... 102

4.23 Discussion for Q3 and Q7 of RT2 ................................................................................. 104

4.24 Summary of results in the survey of calculus ............................................................... 107

4.25 Interview ....................................................................................................................... 108

 4.26 Summary of response to the four research questions: .............................................. 113

Chapter 5 Logarithm ................................................................................................................... 116

5.1 Introduction ....................................................................................................................... 116

5.2 Purpose of the study ..................................................................................................... 116

5.3 Research questions and hypotheses.............................................................................. 116

5.4 Literature Review ......................................................................................................... 117

5.5 Method ......................................................................................................................... 118

5.6 Data collection.............................................................................................................. 121

5.7 Data analysis ................................................................................................................ 123

5.8 Results and discussion.................................................................................................. 123

5.9 Qualitative findings ...................................................................................................... 124

5.10 Interview findings ......................................................................................................... 126

5.11 Discussion and Conclusion ........................................................................................... 127

Chapter 6 Summary, conclusions, and recommendations .......................................................... 130

6.1 Summary ........................................................................................................................ 130

6.2 Limitations ...................................................................................................................... 135

6.3 Implications and recommendations for future research ................................................. 136

6.4 Conclusion ...................................................................................................................... 138

References ................................................................................................................................... 140

List of Figures

Figure 1.a) System of two equations using GC b) Transformation of function using GC ........... 31

Figure 2. a) Transforming function by Desmos b) Graph of a piecewise-defined function ......... 32

Figure 3. a) Graph of a rational function, b) Zeros of the polynomials ........................................ 32

Figure 4. Sample problem from Logarithms app .......................................................................... 33

Figure 5. A Sample problem on apps............................................................................................ 34

Figure 6. Exploring limits of function by table using Desmos ..................................................... 36

Figure 7. Limit with graph using Desmos .................................................................................... 37

Figure 8. a) Derivatives using Desmos b) Instantaneous rate of change of a function................. 37

Figure 9. Example of visualizing definite integral using Desmos ................................................ 38

Figure 10. a) Value of definite integral b) Exploring maximum of a function............................. 39

Figure 11. Bloom’s six levels of learning ..................................................................................... 43

Figure 12. Histogram of the grades of review tests 1 and 2 ......................................................... 47

Figure 13. Grade range of review test 1 and 2 .............................................................................. 48

Figure 14. Grade range of the concept tests .................................................................................. 50

Figure 15. Histograms of concept test grades for treatment and control groups. ......................... 52

Figure 16. Example students work that all categories all complete. ............................................. 58

Figure 17. An example of level two solution................................................................................ 59

Figure 18. An example of level 1 solution.................................................................................... 60

Figure 19. An example of student work for finding domain ........................................................ 60

Figure 20.An example of a student solution with using GC ......................................................... 66

Figure 21. An example of the student solution without using GC................................................ 67

Figure 22. Sara’s paper on RT1 with GC ..................................................................................... 70

Figure 23. Sara’s paper in RT2 without GC ................................................................................. 70

Figure 24. An example of solution with level 3 score by analytic rubric ..................................... 74

Figure 25. An example of solution scored by analytic rubric ....................................................... 75

Figure 26. A solution with logical reasoning and organization of written work scores of 2 ........ 75

Figure 27. Question 3c .................................................................................................................. 76

Figure 28. End behavior of a function .......................................................................................... 76

Figure 29. Question 3g.................................................................................................................. 76

Figure 30. Question 3g with LR=3 but OR= 2 ............................................................................. 77

Figure 31. A sample of work of student in control section on domain problems. ........................ 81

Figure 32. A sample of work of a student in the treatment section on domain problems. ........... 82

Figure 33. Sample of students answer on finding zeros in the treatment section. ........................ 83

Figure 34. Sample of students answer in control section.............................................................. 83

Figure 35. A sample of student answer in control section ............................................................ 84

Figure 36. A sample of student answer in treatment section. ....................................................... 84

Figure 37. Control Pre-SAS .......................................................................................................... 86

Figure 38. Treatment pre-SAS ...................................................................................................... 86

Figure 39. Control post-SAS......................................................................................................... 87

Figure 40. Treatment post-SAS .................................................................................................... 87

Figure 41. A sample of student papers in the treatment section ................................................. 103

Figure 42. A sample of student papers in the control section ..................................................... 103

Figure 43. Sample of students answer for piecewise function.................................................... 104

Figure 44. Sample of students’ answer of limits......................................................................... 105

Figure 45. Sample of students answer in treatment section on definite integral. ....................... 105
Figure 46. Sample of students answer in treatment section on definite integral. ....................... 106

Figure 47. Sample of students answer in control section on definite integral. ........................... 106

Figure 48. A Sample problem on apps........................................................................................ 120

Figure 49. Sample problem from Logarithm app. ...................................................................... 121

List of Tables

Table 1. Sample items of Tharp SAS............................................................................................ 26

Table 2. Grade weights for each portion of the syllabus. ............................................................. 27

Table 3. Sample questions of the review test 1............................................................................. 29

Table 4. Sample questions of the review test 2 ............................................................................. 30

Table 5. Final grade weightings .................................................................................................... 35

Table 6. Sample question of review test 1 .................................................................................... 39

Table 7. Sample questions of review test 2................................................................................... 40

Table 8. A sample of interview questions ..................................................................................... 42

Table 9. Sources that were used to answer each research question .............................................. 46

Table 10. Summary of RT1 (GC- based). ..................................................................................... 47

Table 11. Summary for RT2 (Non- GC based) for all the sections. ............................................ 48

Table 12. Two sample t-test over a mean of RT1 and RT2. ......................................................... 48

Table 13. Two sample t-test over a mean of Q3, 4, of RT1 with Q3, 1 of RT2 respectively. ...... 49

Table 14. Two sample t-test over the mean of the concept test. ................................................... 51

Table 15. Qualitative Rubric (QR) for qualitative analyses. ........................................................ 53

Table 16. Proportional stratified random sampling. ..................................................................... 54

Table 17. Questions of RT1 and RT2. .......................................................................................... 54

Table 18. RT1 by the analytic rubric. ........................................................................................... 61

Table 19. RT2 by the analytic rubric. ........................................................................................... 61

Table 20. t-test over mean scores of the qualitative rubric on review test................................... 62

Table 21. RT1. .............................................................................................................................. 65

Table 22. RT2. .............................................................................................................................. 65

Table 23. Number of students in each stratum. ............................................................................ 74

Table 24. The results of applying the rubric. ................................................................................ 77

Table 25. Sampling information. .................................................................................................. 78

Table 26. The results of applying the rubric. ................................................................................ 78

Table 27. t-test over mean scores of the qualitative rubric on concept test. ................................. 79

Table 28. Concept test control group. ........................................................................................... 80

Table 29. Concept test treatment group. ....................................................................................... 80

Table 30. t-test of means of responses for pre-survey between control and treatment. ................ 89

Table 31. Students’ positive views in control and treatment groups ............................................ 90

Table 32. T-test over means of the scores, for post-survey between control and treatment. ........ 90

Table 33. Students' positive self-impression of their mathematics skill. ...................................... 92

Table 34. Students positive view about mathematics ................................................................... 92

Table 35. results of pre- and post-SAS on the general and personal question.............................. 93

Table 36. Survey Results. ............................................................................................................. 94

Table 37. Students’ responses in pre- and post-survey ................................................................. 98

Table 38. Two sample t-test over the mean of the RT1. ............................................................... 99

Table 39. The result of the t-test over RT2. ................................................................................ 100

Table 40. RT1 control and treatment sections by the analytic rubric. ........................................ 101

Table 41. RT2 control group by the analytic rubric.................................................................... 101

Table 42. RT2 treatment group by the analytic rubric. ............................................................... 102

Table 43. Mean and p-value for the control and treatment sections ........................................... 124

Table 44. Classification of students’ correct responses for Q1. ................................................. 125

Table 45. Categories of students’ responses for Q2. .................................................................. 126

Table 46. Students’ most common mistake ................................................................................ 126
Overview

In this dissertation, the effects of using technology on students’ understanding in calculus

and college algebra were investigated. Chapter 1 is an introduction to the study and a review of

the significance of the study. In addition, chapter 1 describes the purpose of the study, the problems

and research questions. Two theoretical frameworks in the literature that are adapted to using

graphing calculators in teaching mathematics are reviewed at the end of the chapter.

In chapter 2, the literature related to the effects of using technology in different aspects of

learning and teaching mathematics is reviewed. In this chapter, previous studies of using

technology in education and its relationship with theories of learning are discussed. Most of the

research studies on the use of technology in teaching mathematics courses are limited to the use of

a graphing calculator. Therefore, most of the chapter is dedicated to different aspects of the use of

graphing calculators in teaching mathematics courses. These aspects include the effectiveness of

using graphing calculators and the effect of graphing calculators on students’ achievements and

learning skills. In addition, in this chapter some research studies on the use of smartphone and

tablet applications (apps) in mathematical education are reviewed.

Chapter 3 presents the research methodology. In this chapter, different sources of data are

described, including results of review tests and attitude surveys. Samples of review test items and

attitude survey questions are provided as well. In addition, this chapter describes the participants

in the study and samples of methods of teaching with graphing calculators and smartphone and

tablet apps.

Chapter 4 provides a detailed description of data analysis and interpretation of the results

for both college algebra and survey of calculus courses. The first part of the chapter is devoted to

the quantitative and qualitative analysis of data from college algebra and the second part to the

1
survey of calculus. Data were collected and analyzed from different sources including students

pre- and post- attitude surveys, the concept test, review tests 1 and 2, students’ ACT scores, and

student interviews. Summaries of the answers to the research questions are provided at the end of

this chapter.

Chapter 5 i s a report of a research study on the effectiveness of using apps in teaching

logarithms to students in college algebra classes. Data from students’ written tests and interviews

were collected and analyzed qualitatively and quantitatively. Students’ understanding of

logarithms concepts categorized into five levels and students’ errors were identified and

categorized.

Chapter 6 provides a summery, discussions and findings of the research study. In addition,

limitations and recommendations of this research study are discussed.

2
Chapter 1: Introduction

Today, technology has become a significant part of secondary and college mathematics

classrooms. T here is little doubt that appropriate use of technology can enhance learning and

enliven the teaching environment. The general effects of technology on teachers, students, and the

ways that they communicate are undeniable. However, how technology affects teachers, students

and their communications is open to exploration.

Technology has influenced societies extensively in recent years, and its influence is

increasing every year. Graphing calculators have had a broad impact on teaching and learning

mathematics in the past few decades. More recently, teaching and learning of mathematics has

been widely affected by computer software and tablet and smartphone applications (apps).

In 1975, t he National Advisory Committee on M athematical Education suggested that

students in grade eight and above should have access to calculators for all assignments and tests.

In 1980, the National Council of Teachers of Mathematics (NCTM) recommended that students

in all grades take advantage of using calculators in mathematics classrooms.

Several research studies explored the effect of technology on t eachers’ instruction,

students’ learning skills, textbooks, and assessments. M any types of technology have been

produced to improve various aspects of students’ learning: problem-solving, reasoning, and

conceptual understanding. Technology has allowed books and class activities to be designed such

that students have more chances to explore and to visualize mathematical concepts. Assessments

and class activities have changed in today’s technological world. For example, problems that

require long and complicated computations can be included because of using technology in

secondary and college classrooms. The way that students produce and report answers can be

affected by technology as well.

3
 The possibilities of using technology to enhance the efficiency of learning mathematical

concepts pointed research toward investigating the effect of using technology on s tudents’

achievement in college algebra and calculus.

1.1 Purpose of the study

Based on the experience of teaching college algebra for a couple of years and discussion

with more experienced instructors of the course, lack of understanding of some mathematical

concepts by college algebra students was identified. For example, most college algebra students

struggled with the concept of logarithms. In addition, college algebra students had difficulty

making connections between mathematical concepts and representations of these concepts such as

their graphs. Most of the research previously conducted in this area was based on overall students’

performance, which did not identify the difficulty in college algebra students’ understanding of

mathematical concepts. Therefore, this study aimed to tackle this problem by investigating the

effect of using technology (apps and graphing calculators). Graphing calculators and graphing

utilities (for example, Desmos) also were used to help students make better connections between

mathematical concepts and graphical representations. Unlike some previous research about this

study, the written works of students were analyzed.

1.2 Statement of the problem

In college algebra class, students struggle to understand some concepts and mainly rely on

memorizing procedures to solve a problem. Moreover, most of the time, they cannot justify their

answers. College algebra and calculus students have difficulty making connections between

mathematical concepts and their graphical representations. In addition, in both courses, students

have difficulty defining concepts and using procedures to define concepts. Students in college

algebra class have difficulty in understanding logarithm concepts. In this study, two educational

4
apps were used to investigate whether students understand logarithms better by using the apps.

Students used graphing calculators in both courses. However, it was not clear how students use

graphing calculators for solving problems. Therefore, it was important to know how students

approach a problem using technology. For this research, students’ written works were analyzed

to investigate how students used graphing calculators to solve problems and how graphing

calculators affected their understanding. In addition, some students were interviewed about their

approaches to solving problems using technology.

1.3 Research questions

This study was guided by the following research questions.

1. How does the use of technology affect college algebra and calculus students’ understanding

and performance?

2. What areas of college algebra and calculus are affected more by technology?

3. How does using technology affect the organization of college algebra and calculus students’

written work?

4. Does the use of technology positively impact college algebra and calculus students’

attitudes toward their mathematics skills?

1.4 Theoretical framework

The use of a graphing calculator as a tool to introduce and analyze mathematical concepts

has been suggested by NCTM (National Council of Teachers of Mathematics, 2000; National

Council of Teachers of Mathematics. Commission on Standards for School Mathematics, 1989).

NCTM states that graphing calculators transform the classroom into an environment where

students and instructors act as partners in developing mathematical understanding and enhancing

students’ problem-solving skills. This suggestion can also be applied to the use of other

5
technologies such as online graphing calculators and apps in the classroom. In fact, online graphing

calculators and apps not only are as effective as hand-held graphing calculators, but they also have

extra features such as practice problems and are more user friendly than hand held graphing

calculators. In addition, using different technologies in the classroom is aligned with the concept

of multiple and external representation which enables students to implement one representation

and link it to another representation (Goldin & Kaput, 1996).

R.E.Clark (1983, 1991, and 1994), author of several books on t he effect of media on

education, believes that the reason that teaching with one medium is more effective than another

is not the medium itself but the methods that employ the media (Clark, 1994). Clark believes that

media do not have effects on learning beyond being methods to convey information efficiently.

Clark likens the effect of media on learning to the effect of a truck that transports groceries on

people’s nutrition. However, some researchers argue that although there are some students who

learn with or without media, there are some students who will not learn some concepts without

using media. They believe that educational technology is not a natural science, rather, a designed

science (Glaser, 1976; Simon, 1996). Kozma (1994) argues that not observing a relationship

between educational media and learning is because the designer does not make this relationship.

This belief emphasizes the importance of how technology should be implemented to impact

learning. Kozma objects to Clark’s view of separating method from medium and states that

medium and method are integrated parts of design. Medium makes method more powerful and

adds concentration to it.

Clark (1994) has a view that suggests that teachers should not discard the current

technology in favor of new technologies. He suggests that teachers should maximize the effect of

6
current technology in their teaching. Therefore, this study on the effect of a graphing calculator is

aligned with Clark’s suggestion.

Cognitive load theory also has implication on t he use of graphing calculators in the

classroom. This theory, which is based on s everal studies (Cooper & Schleser, 2006; Sweller,

1988; Sweller, Chandler, Tierney & Cooper, 1990; Sweller, 1994; Sweller & Chandler, 1994) ,

states that the learning process has three steps: sensory memory, working memory, and long-term

memory. Everyday, people are faced with a large amount of information that is stored in sensory

memory. A small portion of sensory memory becomes working memory as people start processing

some of this information. As learners practice the working memory and encode it into long-term

memory, the learning process happens. Therefore, the only information that becomes part of

people’s knowledge is long-term memory. This theory holds that cognitive load has two types.

Intrinsic cognitive load is the load that learners utilize to create links between their knowledge and

new knowledge. Extrinsic cognitive load comes from materials - pedagogical tools and methods.

The use of graphing calculators in teaching college algebra can reduce extrinsic cognitive load

imposed on the learners (Chval & Khisty, 2001; Ellington, 2006).

1.5 Significance of the research study

A limited number of research studies have reported on students’ written work in the

presence of technology. Some research studies have compared the overall achievement of students

with technology and students without technology (paper and pencil skills), but no comprehensive

research study has reported on organization and content of students’ written works. This research

study is distinguished for several reasons. First, this research study is one of the first to explore the

effect of hand-held graphing calculators, online graphing utilities (for example, Desmos), and

educational apps on students understanding and organization of written work. The lack of research

7
in this area provides the opportunity to explore the effect of technology on students’ understanding

and organization of written work. A second reason that makes this research study significant is

analyzing students’ papers in addition to analyzing students’ grades. Analyzing students’ papers

helps explore the effect of technology on different aspects of teaching and learning. For example,

how teachers design questions and what kind of answer they expect to observe would be affected

by the presence of technology.

A third reason that distinguishes this research study is analyzing students’ types of errors

when they use technology. Analyzing students’ papers produced in the presence of technology

gives one an opportunity to find students’ errors, thereby affecting teaching and learning. For

example, when teachers become aware of students’ errors, they consider alternate strategies for

teaching the concepts. For example, teachers may use technology to improve visualization of the

concepts, or teachers may present more examples related to the concepts. Teachers’ understanding

of students’ types of errors can affect teacher and student interactions as well.

1.6 Summary

This dissertation reports on e ffects of technology on students’ understanding in college

algebra and calculus classes. In fall 2016, eight college algebra sections with 315 students and in

summer 2017, two calculus sections with 40 students at the University of Arkansas were chosen.

In addition to the effect of using technology on students’ understanding, their organization of

written work also was studied.

8
Chapter 2: Literature review

For several decades, technology has been used as a tool to enhance students’ appreciation

and understanding of mathematical topics. Teachers have used technologies such as computers,

software packages, graphing calculators, educational tablets, and portable devices to facilitate

teaching mathematical concepts.

This chapter provides a review of reports on the results of using technology in secondary

and college mathematics classrooms. Because the focus of the research is on the graphing utilities,

graphing calculator, Desmos, and a few apps will be emphasized. In addition, an overview of the

existing literature as it relates to students’ conceptual understanding of mathematics in the presence

of technology is presented.

2.1 Technology in education

Technology is a combination of tools and processes that facilitate fulfilment of objectives.

Technology has revolutionized every aspect of human life. With the advent of each new

technology, people in different fields evaluate the effect of using the new technology to accomplish

their objectives. Educators also have enthusiastically used the capacity of new technologies in

teaching. They also have attempted to adapt new technologies to theories of learning. Technology

in education has been viewed differently over time, although most view specified learning as the

objective of using technology. In 1963 t he view of technology in education had a focus on

controlling the learning process (Ely, 1963). Januszewski (2001) believes that this view and

specifically use of word control was a result of behaviorism ideas as the dominant theory of

learning at that time. In addition, the notion of managing and controlling learning was widespread

among educators. For example, Hoban (1965) placed the learning-teaching relationship as a part

of learning management. The same idea prevailed in the work of Schewn (1977) and Heinich

9
(1984). They believed that technology should be used as a tool to control teaching and learning.

There were some views that concentrated on di fferent processes such as educational problem

solving and design of educational process. For example, Silber (1970) had a definition that focuses

on the problem-solving skills and not necessarily on increasing the possibility of learning. The

Association for Educational Communications and Technology (1977) had a definition of

educational technology that focused on activities without mentioning learning in the definition.

The current definition of educational technology focuses on t he facilitating role of educational

technology in the learning process. The notion of the current definition is that educational

technology does not cause learning and only has a facilitating role. This means that although

teachers can help learners to learn better, learning is owned by the learners (Robinson, Molenda,

& Rezabek, 2008).

2.2 Technology and theories of learning

Different theories of learning can be adapted to educational technology. Educational

technology can be adapted to the three main theories of learning: behaviorism, cognitivism, and

constructivism.

Behaviorism, developed in the early 20th century, is based on experiments performed on

the learning behavior of animals (Skinner, 1938; Skinner, 1953). Constructivism focuses on the

idea that students should be active in the learning process rather than being passive. This theory

states that knowledge is constructed by students by adapting their existing knowledge to the new

knowledge or creating representations for the new knowledge. Therefore, in constructivism,

students construct the meaning (Piaget & Inhelder, 1969; Vygotski, 1987; Vygotsky, 1980).

The use of technology in education can be consistent with different theories of learning.

Constructivism as the most recent and most dominant theory of learning emphasizes students’

10
exploration. E ducational technology improves learning because students can explore different

concepts and construct cognition of different concepts. Using technology in education increases

the exploratory potential and skills of students. For example, using graphing calculators allows

students to explore how various functions behave. White and others (White-Clark, DiCarlo, &

Gilchriest, 2008) suggest that graphing calculators should be used on a regular basis and this use

should not be restricted to the graphing parts of courses. Graphing calculators or any apps that

provide visual representations of functions are helpful and provide opportunities for students to

explore different kinds of functions. For examples, graphing calculators can be used to find the

points of intersection of two graphs and visualize the concept of solution or roots of an equation.

These examples show how using a graphing calculator can support constructivist learning.

2.3 Overview of different types of graphing calculators

Handheld graphing calculators: A graphing calculator as used here is a handheld

personal computer that can be used to perform calculations, plot graphs, and solve equations. In

addition, there are many smartphone applications and websites that have the same and additional

abilities. In 1985 Casio introduced the first commercial graphing calculator, the fx-700G. Many

other companies have since produced graphing calculators with different features and abilities.

Sharp produced its first graphing calculator in 1986, HP in 1988, and Texas Instrument in 1990.

Most graphing calculators have functionality other than calculation and graphing. Examples of these

functionalities are performing matrix algebra, computing statistics and describing distributions,

finding roots, and evaluating symbolic derivatives and integrals. Some graphing calculators

perform parametric algebra and find antiderivatives as well as derivatives of a given function. The

most recent version of graphing calculators made by TI is TI-Nspire CX. This version has

improved graphical representation and can show multiple views and exhibit graphical animations.

11
 Online graphing calculators: There are numerous websites and smartphone applications

that have the same and additional abilities as handheld graphing calculators. Some of these online

utilities include other options such as mathematics practice at different levels. Desmos is one of

the more popular online graphing utilities. Desmos is available as a website and as an app.

Mathway is a similar online graphing utility that has many extra options with the ability to plot,

solve for a parameter, and explore trigonometry and linear algebra. Mathway solves equations with

intermediate explanations that play a tutorial role. Symbolab is also a user-friendly online graphing

utility that is very similar to Desmos. There are many practice problems on Symbolab that allow

students to practice in different areas such as algebra and calculus. There are many other online

graphing utilities that share similar functionality.

2.4 Research on the effectiveness of using graphing calculators

Considerable research has assessed the effectiveness of using graphing calculators on

different levels of education and on various mathematical topics. Four types of research on the

effectiveness of using graphing calculators are reviewed here. These four types are students’

overall achievement in mathematics classes, students’ mathematical learning skills, students’

performance in college algebra and calculus classes, and students’ mathematical written work.

Examples of mathematical learning skills are conceptual understanding, visualizing, problem-

solving, and reasoning.

2.5 The effect of graphing calculators on students’ achievement

Several research studies have been conducted to explore the effect of graphing calculators

on students’ overall performance in both college and secondary school mathematics. Harvey

(1993) analyzed data from 55 schools by comparing the mean scores of a “calculus readiness (CR)”

test using graphing calculators and computers in pre-calculus. He considered 22 schools as control

12
sections in which students were taught by traditional methods and the rest of the schools as

treatment sections where students were taught using graphing calculators. This study found

statistically significant positive increases in the treatment schools on the CR test scores.

In another effort, Quesada and Maxwell (1994) compared the performance of pre-calculus

students who used graphing calculators and a textbook that requires graphing tools to students who

were taught by traditional methods, regular textbook and scientific calculators. Quesada and

Maxwell repeated their study through three semesters. T he results of Quesada and Maxwell’s

research showed that students who were taught using graphing calculators in pre-calculus classes

had significantly higher grades on the comprehensive common final exams than students who were

taught by traditional methods.

Dunham and Dicks’ (1994) review of research studies indicated that students were more

active and more involved in group activities in a cl assroom where students used graphing

calculators. In another research study Heller and others ( 2005) explored a relationship between

the use of graphing calculators and students’ achievement in algebra 1. This research was

performed on 458 high-school students in suburban Oregon and Kansas. Students in all classes

used the same textbook and the same final exams. The results of the study indicated that students

who used graphing calculators during class activities indicated higher achievement on the final

exams. F urthermore, the results showed that the scores on the final exams were significantly

higher in the classes where teachers explained how to use graphing calculators for solving a

problem, compared to the other classes.

2.6 The effect of graphing calculators on students’ learning skills

Several research studies have considered the effect of graphing calculators on students’

learning skills such as problem-solving, conceptual understanding, visualizing problems, and

13
reasoning. Dunham and Dick (1994) collected different research studies about the effect of

graphing calculators on students’ learning skills and categorized them into two types, the effect of

graphing calculators on problem-solving and the effect on conceptual understanding. They

mentioned that students who used graphing calculators during class activities and assessments had

better understanding in reading and interpreting graphical information, obtaining more

information, and finding better algebraic representations from the graphs. In addition, students had

better understanding about making connections between graphical, numerical, and algebraic

representations. Dunham and Dick’s review of research indicated that students who used graphing

calculators were more successful at problem-solving, used more flexible approaches, and were

willing to engage and stay longer with a problem. Students also solved more nonroutine problems

when they used graphing calculators.

Another research study on pr oblem-solving (Jones, 2008) that was performed on 46

students in pre-calculus algebra classes at Macon College showed that students did not understand

graphs just because they have graphing calculators. One reason given was that students relied on

graphing calculators for checking basic arithmetic.

Ellington’s (2006) review of research provided some results including: students who used

graphing calculators had a better conceptual understanding of functions, variables, and

applications of algebra; using graphing calculators improved low ability students’ performances;

students who used graphing calculators spent more time in mathematical explorations and problem

solving activities compared to students who did not use graphing calculators ; students were more

likely to use graphing calculators in situations that they thought a graph would help problem

solving processes, but students were less likely to use graphing calculators in situations where

they thought a graph was not required (Ellington, 2003; Ellington, 2006).

14
 Penglase and Arnold’s (1996) research on the impact of graphing calculators in high school

and college mathematics classrooms showed a positive correlation between the development of

visualization skills and students’ mathematical achievement, especially for female students.

2.7 Effect of graphing calculators on students’ performance

Numerous research studies have been conducted in different areas of calculus, mainly

concentrating on the effect of graphing calculators on students’ conceptual and procedural

knowledge. R esearch studies show positive impacts of using graphing calculators on students’

learning of calculus concepts. A review by Hunter (2011) indicates that among varieties of

technology, students benefited most using graphing calculators in a calculus class. Graphing

calculators can be used in different ways for teaching and learning calculus concepts. For instance,

increasing the efficiency of calculations and exploring abstract concepts that are difficult to

visualize, such as end behavior, concavity, differentiability, and continuity.

Porzio (1997) performed a study on t he effect of graphing calculators on students’

understanding of numerical, graphical, and symbolic representations of calculus concepts. This

study was performed on 100 students in three calculus sections. The results show that students

performed better when they used graphing calculators.

In a study on the effect of graphing calculators on c ollege algebra students, Smith and

Shotsberger (1997) used four sections of a college algebra class where two were taught using

classical methods and two with the aid of graphing calculators. This study did not show a

significant difference in achievement between the different sections. However, a significant

difference was observed on genders. In that study, female students’ achievement significantly

increased when using graphing calculators. This fact is also observed in other research studies

(Dunham, 1995; Ruthven, 1990). This fact is possibly due to improvement of students’ confidence

15
by having graphing calculators. Some studies showed the positive effect of graphing calculators

on problem-solving in college calculus courses (Bookman & Friedman, 1994). Some

investigations have reported increased conceptual understanding of students in college calculus

courses (Connors, 1995). In another study (Cunningham, 1991) researchers asked students to use

computer software to perform symbolic manipulation in a first-year calculus course. They found

that students who used the software performed significantly better than the other students on the

same exam. This study also shows that students’ performances did not change when they were

deprived of the software package.

2.8 Students’ written work in the presence of technology

A few research studies have been done on effect of graphing calculators on student written

work. Some researchers compared students learning skills such as problem-solving in the presence

of technology and without technology. Ellington (2003) studied the effect of graphing calculators

on students’ achievement in precollege classes. Students with low or average ability were

investigated separately from students with high ability. Ellington’s study shows that although

graphing calculators had no effect on understanding of mathematical concepts of low or average

ability students, their paper and pencil skills improved. Nevertheless, paper and pencil work of

high ability students did not change when using graphing calculators. In addition, the problem-

solving skills improved when using graphing calculators. A meta-analysis was performed by

Hembree and Dessart (1986, 1992) in which the results of 79 studies were summarized into five

conclusions. One of these conclusions is related to students’ written work. It states that using

graphing calculators simultaneously with traditional instruction not only does not harm students’

paper-and-pencil skills but also improves their paper-and-pencil skills. This conclusion was based

16
on the work on K-12 students. No other study was identified on the effect of graphing calculators

on students’ organization of written work.

2.9 Tablet and smartphone apps in education

Researchers have been investigating the potential of smartphone and tablet applications in

mathematics education. Some researchers have performed case studies following quantitative or

qualitative evaluation of the topic and using different grades and genders, while other researchers

have gone further and tried to build a framework for teaching based on iPad or mobile-learning

(m-learning). The concept of m-learning refers to using portable devices with ability to connect to

the internet, such as tablets and smartphones (Park, 2011). Many studies have been done on

different aspects of using iPad technology in education. Park has compared a learning model that

uses mobile or wireless devices (m-learning) with a learning model that uses electronic technology

(e-learning) for learning and teaching from remote sites. Park used transactional distance theory

to analyze these comparisons (Moore, 1993). This theory defines the significant aspects of distance

learning where the teachers and students are separated. Trying to adopt transactional distance

theory to mobile learning, Park concluded that a new theoretical framework for reviewing mobile

or wireless devices used is required.

Some researchers investigated the acceptance of m-learning among individuals by applying

unified theory of acceptance and technology (UTAUT) (Wang, Wu, & Wang, 2009). They

investigated many hypotheses on effect of various parameters such as performance expectation,

perceived playfulness, effort expectation, social influence, and self-management of learning on

behavioral purpose to use m-learning. The study was done on people from all range of ages and

educational background and used UTAUT successfully by adding two new factors to the theory.

It concluded that all the parameters are important in behavioral intention for using m-learning, for

17
example, the performance expectation and perceived playfulness have great effect on behavioral

intention, while the effect of gender was not visible. Effort expectancy was significant for older

individuals and less significant for the young children

In another approach, Hargis et al. (2014) have presented strengths, weaknesses,

opportunities, threats (SWOT) analysis for mobile learning based on s urveys. This research

showed that students’ engagement in learning is the main strength of m-learning and the weakness

is the teachers’ and parents’ lack of technological skills.

Kearney et al. (2012) tried to build a framework that defines socio-cultural features of m-

learning. They highlighted three features of m-learning: validity, personalization, and

collaboration. Validity refers to the ability of m-learning to contextualize placed learning,

personalization refers to ownership implication of m-learning and learning independently, and

collaboration refers to connections and conversational aspect of m-learning.

In other research, Henderson and Yeow (2012) discussed two theories in education,

behaviorism and constructivism. Behaviorists believe learning happens in a change in behavior.

However, constructivists believe that knowledge is not transferred from teacher to student like

transfer of a physical notion, but children learn by constructing the knowledge themselves rather

than transfer from teacher to students. Henderson and Yeow claimed that only recently educational

technology could accept a constructivist theory. They discussed three main qualities of iPad

learning such as mobility, engagement, and collaboration. O’Malley et al. (2013) produced a case

study to investigate the effect of iPad on basic mathematics learning of disabled students. The

result of this study showed that iPad had a positive effect on learning, students’ interest toward

learning, and students’ engagement.

18
 Another interesting case study was reported by Vakil (2014) on increasing the fluency of

urban youth in creating mobile and tablet apps. Although this work may not be relevant to

mathematics education, the case study concludes that it is essential to teach people how to prepare

their own apps instead of just letting them browse and use pre-made apps. This approach was

designated as increasing digital fluency.

2.10 Mathematical apps in secondary and college courses

By the advent of first smart phone in 2007, the possibility of using apps in education was

evaluated by educational researchers. The use of apps in education seems to be very exciting

because, currently, almost everyone has a smartphone. In addition, apps can be designed for very

specific purposes. Therefore, one can design apps to be used for just college algebra or for use by

disabled students in mathematics. In addition, apps not only can be used in class but can engage

students at other times. Much research has been directed at the effectiveness of apps on secondary

and college mathematics courses.

One of the main problems that teachers face is the variety of apps, which makes the

selection of an appropriate app difficult. In addition, most of the apps are not consistent with the

theories of learning, which makes them not suitable to be used as a part of curriculum. In the light

of this fact, Handal ( 2013) evaluated 100 mathematical educational applications and classified

them into nine groups based on functionality. This effort helps teachers to choose suitable apps

effectively from many available apps. For instance, apps are classified based on characteristics

such as simulators, drawing or graphing, and informative apps. All mathematical educational apps

are categorized into three main frameworks that are called explorative apps, instructive apps and

productive apps. Handal stated that in explorative apps, teachers have a facilitator role and ensure

19
that students do not digress while investigating. Activities that require designing products are in

productive apps category and routine drill and problem-solving activities are in instructive apps.

Moyer-Packenham and her colleagues (2015) observed 100 students from grades 3 to 8

with different apps. This research used a clinical interview and tested performance by pretest apps,

test apps, and posttest apps. Data were analyzed both qualitatively and quantitatively to evaluate

children’s performance and efficiency. The conclusion was that age is an important issue in

learning with apps.

In another study, Falloon (2013) concludes that new technology such as computers in past

years and iPad in recent years have failed to perform up to their potential. Two possible reasons

for this were mentioned. First, there is a lack of mutual appreciation between teachers and

pedagogical models with the potential performance or technology. The second reason is the

potential performance of technology has been made unrealistic, and as various schools prepared

technology indiscriminately, the possible success was not visible. Falloon examined the

interaction of five-year old students with specific apps that were selected by an experienced

teacher. These apps were related to literacy, numeric tasks and problem-solving skills. Rather than

evaluating the performance of students by observation, recorded video of students working with

apps was studied. This study concludes that careful attention is needed for designing the content

of apps. Further, it s ummarizes the importance of apps as providing distracting free,

communicative, prompt feedback. Falloon, finally, doubted the effectiveness of using apps to teach

five-year old students.

Green et al. (2014) emphasized the difficulty of teachers in selecting an appropriate app

for use in their classrooms. A rubric was suggested for selecting apps that rate the apps based on

many features, and if the final grade of the app is high then it is proper for the class. This rubric

20
that is designed for students in 5th to 12thgrade is based on the content, feedback, scientific inquiry,

and navigation.

While most researchers have tried to investigate the effectiveness of apps on learning, Blair

(2013) has summarized basic principles for creating a conceptual mathematical app as follows:

• Apps should be designed based on simplicity.

• Apps should be designed based on specific needs.

• A meaningful educational model should be considered for designing an app. A

good educational model allows students to develop fluency in the topics.

• Apps should support individual needs. The apps should be specifically designed to

be self-paced, success oriented, user controllable, and produce low stress.

• Apps should support teachers and parents; these supports can be in terms of extra

comments that suggest the prerequisites of an activity or application or any

direction about implementation of apps in the classroom.

The potential use of apps in education is known to educators. However, teachers experience two

challenges in using apps in class activities. The first challenge is the number of available apps that

makes it difficult to choose among them. The second challenge is that most of the apps that are

built as educational apps have financial purposes and more likely are not based on t heories of

learning. Therefore, it is essential that educators check the consistency of apps before

recommending for use in the classrooms.

2.11 Summary

In summary, educational technology is defined as tools and procedures to facilitate

learning. For a long time, these technologies were personal computers and graphing calculators.

Large volumes of research exist on the effectiveness of graphing calculators. These studies were

21
on all grades from 1st grade to college courses. Most of the studies showed a significant difference

in students’ achievement and learning when using graphing calculators. Therefore, based on the

literature, it is known that graphing calculators can be effective, but it is important to know how

they should be used. In addition, a few research studies have been performed on the effect of

graphing calculators on students’ written work.

The use of apps as educational technology has been investigated by many researchers.

Some of these studies are fundamental studies that build a framework for using apps as educational

technology, and some others are case studies. Most of these studies concluded that apps are

effective in teaching mathematics courses.

22
Chapter 3 Methodology

This chapter presents the details of analyzing qualitative and quantitative data to examine

the effect of technology on the written work of college algebra and survey of calculus students in

fall 2016 and summer 2017 at the University of Arkansas. The instruments that were used to collect

and evaluate data are described.

College algebra: College algebra was a three-semester credit hours course. During the fall and

spring semester students participate in three, fifty-minute class meetings every week. College

algebra students were mostly freshman. College algebra topics were: functions and transformation

of functions, linear and quadratic equations, exponential and logarithmic functions, and system of

linear and nonlinear equations.

Survey of calculus: Survey of calculus was a three-semester credit hours course in polynomial

calculus. During the fall and spring semester students participate in three, fifty-minute class

meetings every week. Survey of calculus was taken by business students or students who do not

plan to take further calculus courses. Survey of calculus concepts were: limits, derivatives, and

definite and indefinite integrals.

3.1 Data collection instruments

The primary goals of the study were to explore the effects of using technology, such as

graphing calculators and apps, on students’ organization of written work, and students’

understanding of specific concepts. Examples of these concepts are logarithms in college algebra

and limits, derivatives, and integrals in survey of calculus. Therefore, to discover the effects, it

was essential to assess students’ written work as fully as possible. To probe students’ abilities to

answer questions properly and to evaluate students’ written work in the presence of technology,

students’ explanatory answers were needed. However, students used the MyLabsPlus (hereafter,

23
MLP) website for all their assignments and tests for both college algebra and calculus courses;

therefore, students’ written works were not accessible.

MLP questions were mostly multiple-choice items with four choices or short answers, but

evidence was needed to show how students used technology, specifically the graphing calculators,

to produce an answer, and how this technology affected their work. Therefore, the current MLP

exams were not sufficient for evaluating; other tools were required. After reviewing related

literature, especially, chapter 17 of Fey, et al. (Fey, Cuoco, Kieran, McMullin, & Zbiek, 2003) two

written tests were designed. Review test 1 and review test 2 were constructed and included open-

ended questions to determine the effect of the graphing calculators on different aspects of students’

written work, such as organization, reasoning, and ability to solve a problem. Details of

participants, the review tests, and course content follow below. Another test, called the concept

test, was part of the regular class. A sample of the concept test, review test 1 and review test 2

questions are included in appendix A, and complete details of the results and analysis are provided

in chapter 4.

3.2 Participants

In fall 2016, eight college algebra sections with a total of 315 students and in the summer

of 2017, two sections of survey of calculus class with a total of 40 students were chosen. College

algebra students were freshmen; calculus students were a mix of freshmen, sophomores, and

juniors. Students in both college algebra and calculus comprised various majors, and races, and

both genders. College algebra was taught by two female and two male instructors, three of them

international and all senior PhD students with similar experience in teaching college algebra

courses. Each instructor taught one control and one treatment section. All instructors had

24
experience in teaching other undergraduate courses. The principal researcher and one of the

instructors taught calculus in summer 2017.

3.3 Student attitude survey

A previously validated student attitude survey (SAS) meeting certain requirements was

needed to examine students’ attitudes toward technology in mathematics courses. The student

attitude survey should:

• Have been previously implemented in other studies.

• Have been administrated to algebra students.

• Have been used in the USA.

• Have been used at the college level.

• Evaluate students’ attitude toward using technology on mathematics problems.

In search of a survey meeting all criteria, different reports in the literature were reviewed.

One example was a r esearch study that was conducted by Smith and Shotsberger (1997) on

assessing the use of the TI-82 graphing calculators in college algebra. Even though they applied a

student attitude survey to college algebra students, the questions were mostly about students’

attitude toward mathematics, not about using technology for solving problems. Therefore, other

attitude surveys such as one by Korey, Brookstein et al. (2011), and Tharp (1999) were reviewed.

Some of the surveys reviewed were used to evaluate high school students’ attitudes about using

technology in mathematics courses. A few surveys focused on measuring college students’ attitude

toward technology in mathematics (Brookstein et al., 2011). However, these were mostly

administered to college students outside of the USA.

Finally, a survey that was developed by Tharp (1992) and was used by Merriweather and

Tharp (1999) for their research study was selected. This survey was designed to investigate the

25
effect of use of the TI-82 graphing calculators on eighth grade general mathematics students, but

it has not been used at the college level. The chosen survey had 23 items with five levels of

response: strongly agree (SA), agree (A), neutral or undecided (N/U), disagree (D), and strongly

disagree (SD). One item was added to the student attitude survey to explore students’ proficiency

in using a graphing calculator, rated from 0 to 10. Sample items of Tharps’ student attitude survey

are shown in Table 1.

Table 1. Sample items of Tharp SAS

Sample Questions SA A N/U D SD

A graphing calculator can be used as a tool to solve problems

I could not solve before.
I would try harder in math if I had a graphing calculator.
I would do better in math if I could use a graphing calculator.
Since I have a graphing calculator, I do not need to learn to
make graphs by hand.

3.4 Supplementary information on college algebra course content

College students participated in three, fifty-minute class meetings every week. Instructors

taught following the book by Blitzer, Mayne and Pietro (2004). The electronic version of the

textbook was included in MLP. MLP is a computer provided system that students use for quizzes,

exams, final exams, and projects. MLP is used in several undergraduate mathematics courses

including college algebra and survey of calculus. Table 2 shows grade weighting for each section

of class activities and tests. In addition to materials of Table 2, grades of review test 1 and review

test 2 were considered as a replacement for the two lowest grade quizzes.

26
 Table 2.Grade weights for each portion of the syllabus.
Percentage of the total
Homework 8%
Quizzes 15%
Wiki projects 8%
Test 1 12%
Test 2 12%
Test 3 12%
Test 4 8%
Participation 5%
Final test 20%

Students could use graphing calculators approved for use on ACT tests. For examples, the

TI-83 and TI-84 were approved for all quizzes, homework, class activities, and MLP tests.

Instructors used both traditional lecturing and PowerPoint slides to teach lessons. In addition, they

used online graphing calculators or smartphone apps in some lessons. Worksheets were also used

as a class activity for each lesson. Students were required to complete worksheets. The sections

taught earlier in the day were considered control sections in which students were not allowed to

use a graphing calculator on t he concept tests, and those sections taught later in the day were

considered treatment sections in which students were allowed to use a graphing calculator on the

concept tests. The detailed information about the concept tests and review test 1 and review test 2

are presented below.

3.5 Concept test, review tests 1 and 2

The concept test: The concept test was the only paper test normally administered in

college algebra. There were seven open-ended questions, designed by the coordinator of the

course. The topics on the concept test were functions and combinations of functions,

transformation of a graph, linear, quadratic and polynomial equations, exponential and logarithms

functions, linear and nonlinear system of equations, and inequalities. S tudents completed the

concept test two weeks before the final exam.

27
 After negotiating with the coordinator of college algebra, two questions of the concept test

were changed such that students had more opportunities to use graphing calculators for answers.

However, questions were graphing calculator neutral, that is, could be answered with or without a

graphing calculator; therefore, it was unclear whether students would use graphing calculators.

Since the primary goal of the study was to explore students’ understanding of the concepts, the

way that they produce and record answers in the presence of technology, some tools to measure

their knowledge were needed. Using the concept test as the only source of students’ written work

for college algebra was not enough to aid the principal researcher in answering all research

questions. Therefore, precise evidence was needed to address these questions. For this reason, two

written tests were designed.

Review test 1: Two written tests, review test 1 and review test 2, were designed to give

the principal researcher more data to answer the research questions. The review test 1 was a

graphing calculator-based test in which questions were designed in a way that students had to use

a graphing calculator to produce an answer. Review test 1 included five open-ended problems that

were designed to examine students’ understanding of zeros and domains of rational functions, x-

and y-intercepts, holes, turning points, end behavior of functions, and asymptotes. Review test 1

was completed by students in both the control and treatment sections as a review for the midterm

exam. Some questions of the review test 1 are shown in the Table 3.

Review test 2: Review test 2 was a non-graphing calculator-based test that had five open-ended

problems. The problems were designed to measure students’ abilities to derive information from

the graph of a function and use that information to solve a problem. These problems were used to

measure students’ understanding of the mathematical concepts. Therefore, for most of the

questions in the review in the review test 2, graphs were given.

28
Table 3. Sample questions of the review test 1
Q3 Graph the polynomial 𝒑𝒑(𝒙𝒙) = 𝟑𝟑𝒙𝒙𝟒𝟒 + 𝟗𝟗𝒙𝒙𝟑𝟑 − 𝟑𝟑𝒙𝒙𝟐𝟐 − 𝟗𝟗𝒙𝒙 + 𝟔𝟔 and answer the
following:
a) Label the zeros of the polynomial on the graph (with exact values).
b) Label the turning points.
c) Describe the end behavior.
Q4 Let 𝒇𝒇(𝒙𝒙) =
�𝒙𝒙𝟐𝟐 −𝟒𝟒�
be a rational function. Find the following:
�𝒙𝒙𝟐𝟐 +𝟓𝟓𝒙𝒙+𝟔𝟔�
a) x- and y-intercepts.
b) vertical and horizontal asymptotes.
c) holes.
d) domain.
Q5 Find the value of x that solves
𝐥𝐥𝐥𝐥𝐥𝐥 𝟑𝟑 𝟐𝟐𝟐𝟐 − 𝐥𝐥𝐥𝐥𝐥𝐥 𝟑𝟑 (𝒙𝒙 − 𝟑𝟑) = 𝟏𝟏

The review test 2 examined students’ understanding of similar concepts as the review test 1

and was given as a review for the final exam. Some common concepts were placed on both the

review test 1 and the review test 2 to allow the principal researcher to compare students’

understanding of a concept with and without graphing calculators. Comparing student

performances on review test 1 and review test 2 ways aimed at finding answers to the research

questions. Students in the control and treatment sections were not allowed to use graphing

calculators during the review test 2. Examples of the review test 2 questions are shown in Table

4.

3.6 Teaching college algebra with the aid of technology

Different technology such as a h and-held graphing calculator, online graphing utility

(Desmos), and apps were used to help students understanding of mathematical concepts in college

algebra.

29
Table 4. Sample questions of the review test 2
Q1 Let 𝑓𝑓(𝑥𝑥) =
2𝑥𝑥 2 +12𝑥𝑥−14
𝑥𝑥 2 +2𝑥𝑥−15
be a rational function. Answer the following questions and explain how
you use graphing calculators to answer each part.
a) y-int:
b) x-int:
c) vertical asymptote(s):
d) horizontal asymptote:
e) domain:
f) hole:
Q3 Use the graph of the polynomial at the right to answer the following:

What is the minimum degree of polynomial? Why?

What is the sign of leading coefficient? Why?
What are the zeros and local max/min for this polynomial?
Use the information you can gather from the graph to write the equation of the
polynomial.

Q5a Solve the following equation:

log6 (x + 4) + log6 (x + 3) = 1

TI-84 in college algebra: Instructors taught students to use a graphing calculator for different

concepts such as domain and range of functions, function transformations, and solving system of

equations. Figure 1a shows the graph of system of two equations and their intersections. Figure 1b

shows using graphing calculators to find the transformation of function.

30
 Desmos in college algebra: Instructors used the online graphing utility Desmos to help

students’ graphical understanding of the college algebra concepts such as transformation of

functions, zeros of functions, piecewise defined functions, and linear and quadratic equations.

Figure 1.a) System of two equations using GC b) Transformation of function using GC

Students can use Desmos to visualize the effect of horizontal and vertical transformations. They

also can easily observe how function’s graph can be stretched or compressed horizontally or

vertically. Samples of transformation of function 𝑓𝑓(𝑥𝑥) = 𝑥𝑥 2 are illustrated in Figure 2a. In

addition, students can observe the behavior of a piecewise-defined function and find the value of

the function at any point. For example, Figure 2b shows the graph of piecewise-defined function

𝑥𝑥 2 − 1 𝑥𝑥 ≤ 0
defined below. 𝑓𝑓(𝑥𝑥) = �
√𝑥𝑥 + 5 𝑥𝑥 > 0

With the use of graphing calculators students can make connection between graph of a

31
 a) b)

Figure 2. a) Transforming function by Desmos b) Graph of a piecewise-defined function

function and algebraic definition of asymptotes, holes, and end behavior of the function. Figure 3a

shows the graph of a rational function. Using Desmos, students can visualize zeros, multiplicity of

zeros and turning points of graphs of functions. Figure 3b s hows zeros and turning points of

polynomial 𝑝𝑝(𝑥𝑥) = 6𝑥𝑥 3 + 9𝑥𝑥 2 − 6𝑥𝑥.

a)
b) a)

Figure 3. a) Graph of a rational function, b) Zeros of the polynomials

32
 Apps in college algebra: Logarithm is an important concept in mathematics, and anecdotal

evidence from instructors has confirmed that students in college algebra have difficulties

understanding and using logarithms. Instructors used smartphone applications for teaching

logarithms. Instead of worksheet activities, students were asked to complete activities in the pre-

specified apps of “Logtrainer” and “Logarithms.” Students were also asked to send screenshots of

their results to their instructors. The Logtrainer app was used at the first session. The Logtrainer

app is a t utorial and practice-based app, containing multiple-choice questions. In this app,

questions were mostly about converting logarithms to exponentials. After clicking on one of the

answer choices for a question, learners could see the correct answer and a complete explanation of

a similar problem. In this app, similar problems are repeated several times, supposedly to help

learners understand logarithms using their knowledge of exponentials. A sample problem from

this app is shown in Figure 4.

Figure 4. Sample problem from Logarithms app

33
 Figure 5. A Sample problem on apps

The Logarithms app consists of four parts: logarithm rules, simplification, expansion, and

solving logarithmic equations. Each part includes some multiple-choice problems relating to the

subject (examples are shown in Figure 5.). Students were able to see correct answers immediately

after picking one of the answer options. The complete solution to each question was provided in

this app; therefore, students were able to review their work and correct themselves.

Survey of calculus

After working on college algebra, the principal researcher decided to examine the same

questions on the more advanced course, survey of calculus. Using technology might be more

efficient for understanding concepts such as limit, derivative, and integral. Therefore, a similar

study was conducted to examine how using technology, specifically a graphing calculator, would

affect students’ understanding of mathematical concepts and organization of written work. In the

summer of 2017, t wo survey of calculus classes with a total of 40 students were chosen. The

section that was taught by the principal researcher was considered the control section and the

34
section that was taught by another instructor was considered the treatment section. Students in

both sections were allowed to use ACT-approved graphing calculators, such as the TI-83 or TI-

84, for doing homework, class activities, and all assessments except review test 1 and the review

test 2 for this course. Clicker quizzes as developmental evaluations were included in the

PowerPoint slides presented in class. Students worked on their quizzes, exams, and final exams on

the MLP website.

3.7 Supplementary information of survey on calculus

In the summer section, survey of calculus was a f ive-week course meeting five ninety-

minute periods a week. Instructors taught calculus following the book Calculus with Application

(Lial, Greenwell, & Ritchey, 2013). The e-version of the textbook was included on the MLP

website. The final grades were evaluated based on students’ performance in homework, quizzes,

and tests. The grade weightings are shown in Table 5.

Table 5. Final grade weightings

Percentage of the total
Homework 12%
Quizzes 10%
Test 1 16%
Test 2 16%
Test 3 16%
Clicker quiz and review test 10%
Final test 20%

All the resource materials including the textbook, PowerPoint slides, worksheets, quizzes,

and tests were the same for both sections. Instructors used PowerPoint slides for teaching

mathematical concepts. They also used Desmos to help students with graphical understanding of

the concepts.

35
 3.8 Teaching calculus with the aid of a graphing calculator and Desmos

Instructors used hand-held graphing calculators and Desmos to promote students’

understanding of mathematical concepts such as limit, derivative, and definite integral. Examples

of teaching with hand-held graphing calculators and Desmos are shown in Figure 6.

Figure 6. Exploring limits of function by table using Desmos

Survey of calculus with Desmos: Instructors used Desmos for sketching graphs and visualizing

discontinuities of a function. Figure 6 shows how Desmos can visualize the right side and left side

limit of a piecewise-defined function at x= -1. Students used tables and graphs to investigate limits.

An example of investigating a limit problem using a table is shown in Figure 7. Students plug in

different values for x and explore its effect on y and observe the behavior of the function around a

point of discontinuity (in this example, around x= 0). Other topics taught using technology were

local and absolute extrema, graph of the first and second derivative of a function, and relations of

36
derivatives to increasing/decreasing, concave upward/downward. Figure 8a shows the graphs of

function 𝑓𝑓(𝑥𝑥) = 3𝑥𝑥 3 − 𝑥𝑥 2 − 4𝑥𝑥. The graphs of first and second derivatives are also shown.

Figure 8b indicates instantaneous rate of change of function g(x) = ln(x 2 + 1).

b)

Figure 7. Limit with graph using Desmos

a) b)

Figure 8. a) Derivatives using Desmos b) Instantaneous rate of change of a function

Instructors used Desmos to visualize boundaries of areas between two curves to help students have

a better understanding of the definite integral. Figure 9a illustrates the value of the integral of

37
functions 𝑓𝑓(𝑥𝑥) = 𝑥𝑥 2 − 12, 𝑔𝑔(𝑥𝑥) = 11𝑥𝑥, 𝑥𝑥 = −2, 𝑎𝑎𝑎𝑎𝑎𝑎 𝑥𝑥 = 2, and Figure 9b shows examples of

integral for functions 𝑓𝑓(𝑥𝑥) = 𝑥𝑥 2 , 𝑎𝑎𝑎𝑎𝑎𝑎 𝑔𝑔(𝑥𝑥) = 𝑥𝑥 + 20 by Desmos.

a) b)

Figure 9. Example of visualizing definite integral using Desmos

TI-84 in a survey of calculus: Instructors taught students how to use a graphing calculator

(TI-84) for solving various problems. For example, students could calculate the value of a definite

integral with the TI-84. Figure 10a shows an example of solving definite integral

5 2
𝑓𝑓(𝑥𝑥) = ∫1 �(5𝑥𝑥+1)3 � 𝑑𝑑𝑑𝑑 with the TI-84.

Data collection instruments for calculus

Calculus was an MLP course; therefore, students’ paper tests were not available. Since the

goal of this research was to evaluate students’ mathematical understanding and their organization

of written work using technology, paper tests were needed. Therefore, like college algebra, in

addition to MLP tests, two review tests were designed.

Review test 1 for survey of calculus students: Review test 1 was the first paper test in

calculus that covered the concept of limits and derivatives. This test was administered before the

midterm exam and had four open-ended questions. Table 6 shows an example of the review test 1

38
questions.

Figure 10. a) Value of definite integral b) Exploring maximum of a function

Table 6. Sample question of review test 1

Q1 Find the open interval where the function 𝑓𝑓(𝑥𝑥) = −2𝑥𝑥 3 + 12𝑥𝑥 2 + 170𝑥𝑥 − 6 is
concave upward or concave downward. Find any inflection point.
Q2
a) Let f(x,y) be a function that has (6,7) as a critical point. We determine that
𝑓𝑓𝑥𝑥𝑥𝑥 (6,7) = −2 , 𝑓𝑓𝑦𝑦𝑦𝑦 (6,7) = 2 𝑎𝑎𝑎𝑎𝑎𝑎 𝑓𝑓𝑥𝑥𝑥𝑥 (6,7) = −10
What D test tells us about the function f?
𝜕𝜕𝜕𝜕
b) Find the partial derivative of 𝑧𝑧 = 8𝑥𝑥 + 7𝑥𝑥𝑦𝑦 3 − 6𝑦𝑦 2
𝜕𝜕𝜕𝜕

Q3 Find all the local maxima, local minima, and saddle points of the given function:
𝑓𝑓(𝑥𝑥, 𝑦𝑦) = 4𝑥𝑥 2 + 6𝑥𝑥𝑥𝑥 + 8𝑦𝑦 2 + 4𝑥𝑥 − 20𝑦𝑦

Review test 2 for calculus students: Review test 2 was a comprehensive paper test over

the concepts of limits, derivatives, and integrals. This test was given just before the final exam.

Some questions from review test 2 are shown in Table 7.

3.9 Primary data collection

Data from multiple sources were collected to address the research questions. Some

of these sources were standardized tests that were taken before students began attending college

and some of them were tests designed by coordinators that were part of the material for the courses.

39
Table 7. Sample questions of review test 2.
Q3 a) Sketch a graph for the below function. (10 points)
3 𝑥𝑥 < 0
⎧
⎪
f(x)= 𝑥𝑥 2 + 1 0 ≤ 𝑥𝑥 ≤ 3
⎨
⎪
⎩ 10 𝑥𝑥 > 3
b) Find all values of x where the function f is discontinuous. (Show all steps of your work).
c) For which x value in the interval [0,3] limit of f(x) exists? Why?
Q4 a) Find the average rate of change of the function y over the given points. (10points)
y=�(7𝑥𝑥 + 7) between x=0 and x=6
b) Find the slope and the equation of the tangent line to the graph of the function f(x) at
the given value of x. (show all steps of your work).
f(x)=𝑥𝑥 2 (3 − 𝑥𝑥) ; x=-2
Q5 The graph of 𝑓𝑓 ′ (𝑥𝑥) is given below. Determine which of the following graphs is an

approximate sketch of the f(x). Explain your reason. (5 points)

a b f(x)

c d
Q7 Find the area between the following curves. (Show your work). (5 points)

X=-4, x=3, y=0, and 𝑦𝑦 = 2𝑥𝑥 2 + 4

The following are all the data sources of this study.

Data from student attitude surveys: Meetings were arranged with the coordinators and

instructors of college algebra and calculus courses a week before college algebra class began in

fall 2016, and a week before calculus class began in summer 2017. These meetings were arranged

to discuss the research purposes and procedures. Instructors who agreed to participate in the

research were asked to send their class rosters to the principal researcher. A four-digit code was

40
assigned to each student in the participating classes. Students used the assigned code to remain

anonymous when they completed pre-surveys and post-surveys.

• Data from ACT scores: ACT scores of college algebra and calculus students were

collected for both control and treatment sections.

• Data from concept test: college algebra students’ concept test scores were

collected for analyzing in college algebra.

• Data from review test 1: college algebra students’ review test 1 scores were

collected for analyzing.

• Data from review test 2: college algebra students’ review test 2 scores were

collected for analyzing

• Data from review test 1: calculus students’ review test 1 scores were collected for

analyzing.

• Data from review test 2: calculus students’ review test 2 scores were collected for

analyzing.

Data from MLP tests 1, 2, 3, and final exam in survey of calculus: In addition to the

review test 1 and review test 2 scores, calculus students’ MLP tests 1, 2, 3, and final exams were

collected and organized in different classifications of the treatment and control section to be

prepared for final analyzing.

Data from student interviews: A week before the final exam, after all the written tests

were taken, students in both control and treatment sections of college algebra and calculus were

interviewed, one-to-one and face-to-face. Students were asked general questions about their

attitude toward using technology in mathematics courses and specific questions about the way they

used the graphing calculators for solving the problems on written tests. In addition, questions were

41
asked to clarify how students understood concepts such as end behavior of a function, limits,

derivatives, and integrals. Some of the interview questions are shown in Table 8.

Table 8. A sample of interview questions

Do you think using graphing calculators can help you to understand the mathematics concepts
better? In which area of this course?
Does using a graphing calculator affect the way that you write a solution? How?
What does the lim 𝑓𝑓(𝑥𝑥) = 7 mean? Explain in your own words.
𝑥𝑥→2

How do you use graphing calculators to solve definite integral?

3.10 Primary evaluations

All test questions were grouped into three categories based on Bloom’s six levels of

learning (Bloom, 1984). These six levels are shown in Figure 11. Questions were categorized based

on the levels of learning needed. These categories are a procedure, application, and conceptual

questions. Questions that need recalling the concepts were categorized in the procedural category.

These questions require limited understanding and mostly need memorizing solution steps.

Questions that require the application of learning are categorized accordingly. To solve application

questions students, need to know and use a combination of rules and definitions. Conceptual

questions are those questions that need a higher level of learning: analyzing, evaluating and

creating.

Evaluation of data began by comparing the students’ performances in each category between the

control and treatment sections. This comparison includes both qualitative and quantitative

differences between students’ performances in the presence of technology. In addition, students’

organizations of written work were evaluated. Since this study was performed on two different

courses, college algebra and survey of calculus, there were some specific subjects that were

considered for each course separately.

42
 Create

Evaluate

Analyze

Apply

Understand

Remember

Figure 11. Bloom’s six levels of learning

College algebra: In college algebra, the numerical scores of students on the concept test,

review test 1 and review test 2 were evaluated and compared between control and treatment

sections. Moreover, each question for both treatment and control sections were compared

separately to find out how or whether students’ performances were affected by using a graphing

calculator. Students’ written work was another aspect of this study. For this purpose, the difference

between students’ written work on the concept test, and review test 1 and review test 2 in the

control and treatment sections in areas such as organization, accuracy, and length of answers were

analyzed. Finally, the students’ graphical understanding of the concepts such as zeros, holes,

turning points and end behavior of functions were compared.

Survey of calculus: In survey of calculus, the numerical scores of students on MLP exams,

review test 1 and review test 2 were evaluated and compared between control and treatment

sections. The students’ paper tests were analyzed as in college algebra. In addition, students’

43
understanding of several concepts such as derivative and integral were analyzed between control

and treatment sections.

This study also investigated how using a graphing calculator affects producing a graph,

evaluating definite integral, writing an equation for a tangent line, and understanding of derivatives

qualitatively. Finally, students’ confidences in solving a problem in the presence of technology

were investigated.

3.11 Summary

This chapter explained the primary sources of data collection for investigating the effect of

technology on s tudents’ performances in college algebra and survey of calculus. For college

algebra, three tests were used: concept test, review test 1 and review test 2. For survey of calculus,

data were collected from review test 1 and review test 2. Data from student attitude surveys were

collected for both college algebra and survey of calculus.

Data from the concept test were used to evaluate students’ performances in the control and

treatment sections quantitatively. In addition, this test was used to evaluate the students’

understanding of mathematics and organization of written work. Review test 1and review test 2

were used to compare each student’s performance in the presence of technology and without using

technology. Further, review test 1and review test 2 were used to perform a qualitative comparison

of each student’s understanding of mathematics and organization of written work with and without

technology. The same analysis was applied to review test 1and review test 2 of survey of calculus

course. In addition, students’ attitude surveys were analyzed.

44
Chapter 4 Data analysis and results

This chapter provides a detailed description of data analysis and interpretation of the results for

both college algebra and survey of calculus courses. This study was conducted in fall 2016 on

eight college algebra sections with a total of 320 students and in summer 2017 on two survey of

calculus classes with a total of 40 students at the University of Arkansas Mathematical Sciences

Department. The primary focus of this study was the effect of technology on college algebra and

calculus students’ understanding. Four college algebra sections were chosen as control groups in

which students were not allowed to use graphing calculators on the concept test, and four classes

were considered as treatment groups in which students could use a graphing calculator on the

concept test. Data were collected from different sources including students pre- and post- attitude

surveys, the concept test, review tests 1 and 2, students’ ACT scores, and student interviews.

This study was guided by the following research questions:

1. How does the use of technology affect college algebra and calculus students’ understandings

and performances?

2. What areas of college algebra and calculus are affected more by technology?

3. How does using technology affect the organization of college algebra and calculus students’

written work?

4. Does the use of technology positively impact college algebra and calculus students’

attitudes toward their mathematics skills?

The stage of data analysis was guided by Corbin and Strauss (2008), Miles, Huberman, and

Saldana (2013), and Thompson (2012). Data were analyzed qualitatively and quantitatively to

answer research questions. The data sources and result are described in the following sections.

How the data sources related to the research questions is shown in the Table 9.

45
Table 9. Sources that were used to answer each research question
Research question Data sources
1. How does using • Student's grades on review tests 1 and 2 and concept
technology affect college test.
algebra and calculus students’ • For understanding, data from student interviews and
performances and analytic rubric.
mathematical understandings?

2. What areas of college • Students grades in review test 1 and 2 and concept test
algebra and calculus are • Data from the analytic rubric
affected more by technology? • Data from interviews

3. How does using technology • Data from analytic rubric on review test 1 and review
affect the organization of test 2 as well as the concept test.
college algebra and calculus • Data from student interviews.
students’ written work?

4 Does the use of technology Data from SAS and interviews

positively impact college
algebra and calculus students’
attitudes toward their
mathematics skills?

This research study was conducted on two different courses. The first part of the chapter is devoted

to the analysis of college algebra quantitatively and qualitatively, and, in the second part, a survey

of calculus will be discussed.

4.1 Quantitative data analysis and result for college algebra

Quantitative result of review tests 1 & 2: As mentioned before, two written tests, review test 1

(hereafter, RT1) and review test 2 (RT2), were designed by the principal researcher. RT1 was a

graphing calculator (hereafter, GC) based test, and students used TI-84 to complete the test. RT1

was completed by students in both the control and treatment sections as a review for the midterm

exam. RT2 was a n on-graphing calculator-based test that had five open-ended problems. RT2

46
examined students’ understanding of concepts similar to those in RT1 and was given as a review

for the final exam in both the control and treatment sections. Students were given 50 minutes to

complete each review test and were informed that the result would be a replacement for their lowest

quiz grades or lowest wiki project grades. Students papers for RT1 and RT2 were graded by the

same rubric and were reviewed by the principal researcher and another instructor of the college

algebra. To pair students result a 4-digit SAC was used. Students’ numerical grades for RT1 (GC-

based) and RT2 (non-GC-based) for all sections were collected. The histograms for RT1 and RT2

grades are shown in Figure 12.

Boxplot and the summary of statistical results for RT1 and RT2 are shown in Figure 13 and the

Figure 12. Histogram of the grades of review tests 1 and 2

Tables 10 and 11.

Table 10. Summary of RT1 (GC- based).

Number of the question in RT1 Q1 Q2 Q3 Q4 Q5 Total
Median 1.5 3 4 3 0 11.5
Mean 1.4 2.73 3.78 2.92 1.12 11.95

47
 Figure 13. Grade range of review test 1 and 2

Table 11. Summary for RT2 (Non- GC based) for all the sections.
Number of the question in RT2 Q1 Q2 Q3 Q4 Q5a Total
Median 3 2 2 3.5 1 11.5
Mean 3.06 1.76 2.58 2.33 1.24 10.97

Both boxplots of students’ grades in the RT1 and RT2 show that there is no outlier in the data.

Students mean score for RT1 (GC based) was 11.95 while the mean score of RT2 (non-GC

based) for these students was 10.97. In other words, students had a better performance when they

used a graphing calculator on the similar mathematical concepts compared to when they did not

use a graphing calculator. The median for both review test scores is 11.5.

In next step independent two-sample t-test with 95 percent confidence intervals on t he

result of RT1(GC) and RT2 (non-GC) were applied and shown in Table 12.

Table 12. Two sample t-test over a mean of RT1 and RT2.
Summary of t-test over Mean of RT1 Mean of RT2 df P-value
RT1&RT2 N=170 N=190
Overall grades 11.95 10.98 340.15 .00245

This result gives additional evidence to propose that the GC (TI-84) used by a student in RT1 had

a significant impact on student performance compared to the performance of the non-GC based

48
test. In other words, p-value = .00245 < α= .05 indicates that there is a significant difference

between overall student grades on RT1 and RT2 with a confidence interval of 95 percent. The

mathematical concepts were similar between question number 3 of RT1 and question number 3 of

RT2. Question number 4 of RT1 and question number 1 of RT2 had the same concepts. Therefore,

two-sample t-test over the mean of these questions was applied to find if there is any significant

difference between student performance on the common concepts with and without technology.

The result of the t-test is shown in the Table 13.

Table 13. Two sample t-test over a mean of Q3, 4, of RT1 with Q3, 1 of RT2 respectively.
Number of the Mean of the Number of the Mean of the df p-value
questions in questions in RT1 questions in RT2 questions in
RT1 (N=170) (N=190) RT2
Q3 3.78 Q3 2.58 329.74 0.00000209
Q4 3.50 Q1 3.06 350.41 0.0187

There is a significant difference between students’ performance in Q3 and and Q4 of RT1

and Q3 and Q1 of RT2 with a significance level of α= .05 and confidence interval of 95 percent.

The p-value is 0.0187, which is less than significance level α= .05. This result gives enough

evidence to support alternative hypotheses. Students showed a better performance on s imilar

concepts when they used graphing calculator compare to their performance without a graphing

calculator.

4.2 Quantitative results of the concept tests

As a reminder for the reader, the concept test was the only paper test normally administered

in college algebra. There were seven open-ended questions, designed by the coordinator of the

course. The topics on the concept test were functions and combinations of functions,

transformation of a graph, linear, quadratic and polynomial equations, exponential and logarithms

functions, linear and nonlinear system of equations, and inequalities. Students completed the

49
concept test two weeks before the final exam. Four college algebra classes were considered as

treatment groups in which students used the TI84 GC to complete the concept test and four control

sections in which students did not use the TI84 GC to complete the concept test. Student papers

for the concept test were graded by all instructors of the college algebra courses (each instructor

one question) with the same rubric designed by the coordinator of college algebra. There were 131

students in the control sections and 135 s tudents in the treatment sections who completed the

concept test. The first phase of statistical analysis of the concept test was started by finding a

statistical summary for control and treatment groups and boxplot for showing the distribution of

data and identifying any outliers which is illustrated in Figure 14.

Figure 14. Grade range of the concept tests

The first boxplot corresponds to students’ grades on the concept test from treatment sections and

as it shows in the plot there is an outlier (a zero). The second plot represents students’ grades on

the treatment sections without outlier (outlier was removed) and the third plot belongs to students’

50
grades of the concept test in the control sections. Here is the statistical summary of the students’

grades on the concept test for control sections and treatment sections.

Statistical summary for control sections. N= 131

Min. 1st Qu Median Mean 3rd Qu Max

26.50 48.00 60.00 59.97 70.00 95.00

Statistical summary for treatment sections. N= 135

Min. 1st Qu. Median Mean 3rd Qu Max

25.00 54.00 64.00 63.21 74.00 92.00

The result of statistical summary between control and treatment sections on the concept test shows

that students in treatment sections have a higher mean (63.21) and median (64) scores compared

to mean (59.97) and median (60) of the control sections. Histograms of the grades of control and

treatment groups are shown in Figure 15. In the next step, the researcher conducted a two-sample

t-test to compare the concept test scores between the control and the treatment groups. The result

of the t-test is shown in the Table 14.

Table 14. Two sample t-test over the mean of the concept test.
t-test over the Mean of the concept Mean of the concepts df P-value
concept test between test for control sections test for treatment
control and treatment N=131 sections
sections N=135

59.97 63.21 259.13 0.06783

With a significance level of α= 0.1 and 90 percent confidence interval, we reject null hypotheses

which says there is no difference between the mean of the control group and mean of treatment

group and accept alternative hypotheses which says “there are true differences between the mean

of students’ grades in the control and treatment groups.”

51
 Figure 15. Histograms of concept test grades for treatment and control groups.

In other words, students who used a GC on the concept test performed better compared to

the students who did not use a GC on the same test with a 90 percent confidence interval and a

significance level of α = 0.1.

4.3 Qualitative data analysis and result for college algebra

The result of RT1 and RT2 by the analytic rubric: To find if the use of technology affects

students’ understanding and organization of written work we used a qualitative rubric. A fter

researching similar studies and reviewing literature the following rubric was designed. This rubric

was used to investigate whether the use of GC influences students’ performance on the following

skills: logical reasoning; organization including intellectual order, written order and use of symbol

and notation; and use of a graph. The rubric was designed and applied to the papers of several

college algebra students after it was revised multiple times. This rubric is shown in Table 15.

52
Table 15. Qualitative Rubric (QR) for qualitative analyses.
Level Logical reasoning
No relevant argument to support
0 conclusion or no relevant
conclusion
Some reasoning to support an
incomplete solution
1 or poor reasoning to support
correct solution or strong reason
for the poor conclusion
Incomplete reasons to
support the complete correct
solution
2
or complete reason to support
correct (or incorrect) incomplete
solution
Complete logic argument to
3 support complete correct
conclusion
Organization including
Intellectual order, written
order, Use of graph
correct use of symbols and
notation.
No relevant
No relevant organization evidence
of use of graph
Some organizations for
incomplete or incorrect use of a graph for
solutions or poor organization an
for a correct solution or well incomplete or
organization for the poor incorrect solution
conclusion
well Organized description,
argument, and notation for
correct (or incorrect)
incomplete conclusion or Relevant use
incomplete argument, of graph in parts
description, and notation for
complete correct (or incorrect)
conclusion
Well organized description, complete use of
argument, and complete graph for
correct notation for correct complete correct
conclusion conclusion

The population size (students who have both RT1 and RT2) was 125, which was too large

to evaluate by the rubric. To save time, and have a p recise analysis, a s ample of students was

selected. The following conditions were considered for data sampling. All students who have

only one of the review tests grades were removed. Our goal was to compare students with

themselves to see if the use of technology influences their answer. Students belong to different

categories including good, medium and poor, and researcher wanted to analyze students’ papers

in all these categories, but the numbers of students in each of these categories were not equal;

therefore, the researcher used proportional stratified random sampling.

53
 4.4 Proportional stratified random sampling

Students were stratified into three groups (good grade, average grade, and low grade) but

the number of students was not equal, which means that if an equal number of students from each

stratum would be selected it would bias the result. We have 3 strata with 21, 70, and 30 population

sizes, and we chose a sampling fraction of 1/5 of population size in each stratum which means 4,

14, and 7 population size respectively. Table 16 shows the information about sampling.

Table 16. Proportional stratified random sampling.

Students Students Qualitative Total number of students in Final sample size
grade /100 grade /20 grade /4 each stratum for N_GC test
80 -100 16-20 A, B N1=21 n1=4
60 -79.99 12-15.9 C, D N2=70 n2=14
0-59.99 0-11.9 F N3=33 n3=7

After stratified random sampling, the researcher collected the students’ papers in RT1

and RT2 for each class into separate folders. Some questions were common between RT1 and

RT2. B y comparing the equivalent problem solved with the same student with and without

technology, we might have a more precise view about the influence of technology on students

learning mathematics. Therefore, the rubric was applied to questions with common concepts in

RT1 and RT2. The questions were listed in the Table 17.

Each question has several parts and they were picked from different levels of learning. The

solution rubric for the selected question from the RT1 and RT2 are shown in the following rubric.

Table 17. Questions of RT1 and RT2.

Question numbers in RT 1 Question numbers in RT2
3a 3c(z) (first part)
3b 3c( M)
3c 3b
4d 1e

54
Questions

Q3. Graph the polynomial 𝒑𝒑(𝒙𝒙) = −𝟑𝟑𝒙𝒙𝟒𝟒 + 𝟗𝟗𝒙𝒙𝟑𝟑 − 𝟑𝟑𝒙𝒙𝟐𝟐 − 𝟗𝟗𝒙𝒙 + 𝟔𝟔 and answer the following:
a. Label the zeros of the polynomial on the graph (with exact values).
b. Label the zeros of the polynomial on the graph (with exact values).
c. Describe the end behavior
�𝒙𝒙𝟐𝟐 −𝟒𝟒�
Q4. Let 𝒇𝒇(𝒙𝒙) = �𝒙𝒙𝟐𝟐 +𝟓𝟓𝒙𝒙+𝟔𝟔� be a rational function. Find the following:

e. Domain

Solutions Point allocated

Q3_a:
2: 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑎𝑎𝑎𝑎𝑎𝑎 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑥𝑥 𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣
A= (-1, 0), B= (1, 0), C= (2, 0) 3: �
1: 𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 𝑜𝑜𝑜𝑜 𝑡𝑡ℎ𝑒𝑒 𝑔𝑔𝑔𝑔𝑔𝑔𝑔𝑔ℎ

Q3_b:
1: correct and exact x valus
A= (-.443, 8.5) 3:� correct and exact y value
1:
1: labaling on the graph
B= (1.693, 1.191)
C= (1, 0)

Q3_solution1: As x → ±∞, P(x) → −∞ or as x goes

2: correct explanation and reasoning
3: �
1: correct notation
toward infinity from the right and left, p(x) fell.
Solution 2: since polynomial has degree 4
and leading confection is positive therefore p(x) goes
toward negative infinity as x goes toward
either positive or negative infinity.

Q4_e:
1.5: finding VA and hole
Solution 1: 3:�
1.5: corect answer and notation
Domain= (-∞, -3)∪(-3, -2) ∪(-2,∞)
(𝑥𝑥 − 2)(𝑥𝑥 + 2)
𝑓𝑓(𝑥𝑥) =
(𝑥𝑥 + 3)(𝑥𝑥 + 2)

55
First find hole if it exists: x= -2,
Then find VA, in this case, x=-3, then remove hole and VA from all real number.
Or by using graph

Table: Solution Rubric _RT2 (Non-GC)

Questions
2𝑥𝑥 2 +12𝑥𝑥−14
Q1. Let 𝑓𝑓(𝑥𝑥) = 𝑥𝑥 2 +2𝑥𝑥−15
be a rational function. Answer the following questions and explain how
you find the answer for each part.
e) domain

Q3. Use the graph of the polynomial at the right to answer the following:

b) What is the sign of the leading coefficient? Why?

c) First part) What are the zeros for this polynomial? (second part of c) What is local max/min
for this polynomial?

Solutions Point allocated

2(x+7)(x−1) 1.5: 𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓𝑓 𝑉𝑉𝑉𝑉 𝑎𝑎𝑎𝑎𝑎𝑎 ℎ𝑜𝑜𝑜𝑜𝑜𝑜

Q1: f(x)= 3:�
(𝑥𝑥+5)(𝑥𝑥−3) 1.5: 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 𝑎𝑎𝑎𝑎𝑎𝑎 𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛
VA: x= -5 x=3
Domain:
(-∞, -5)∪ (-5,3)∪(3, ∞)

Q3_b: This graph is a polynomial with odd degree.

2: 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑎𝑎𝑎𝑎𝑎𝑎 𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟𝑟
3: �
1: 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛
The end behaver of polynomial as
x → + ∞ p(x)→+∞ and
as x → −∞ p(x)→-∞ which means the leading
coefficient is positive.

56

Q3_c_ first part: (0,0), (1,0), (2,0), (3,0)
Or x= 0, x=1, x=2, x=3
Q3_c_secod part: Approximately,
local max: (0,0), (1.5, 1)
Local min: (.5, -.5), (2.5, -2.5)
2: 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑎𝑎𝑎𝑎𝑎𝑎 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑥𝑥 𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣
3:�
1: 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛𝑛
1: 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑥𝑥 𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣
3:� 1: 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑦𝑦 𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣𝑣
1: 𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 𝑜𝑜𝑜𝑜 𝑡𝑡ℎ𝑒𝑒 𝑔𝑔𝑔𝑔𝑔𝑔𝑔𝑔ℎ

4.5 Validating the analytic rubric

For re-labeling and validating the qualitative rubric, the analytic rubric was applied to 20

papers and the rubric was revised based on different types of solutions that appeared. In addition,

the researcher asked another colleague to use the rubric for some papers in RT1 and RT2 and asked

him to explain how he scored. He mentioned that he did not have difficulty with scoring

organization and the use of a graph, and it covers all possible type of students’ answers. For scoring

level zero and three of organization, there was not any ambiguity for him as well. He mentioned

that when he wanted to score logical reasoning if students’ logical reasoning was not either

irrelevant or complete he had difficulty to score between level one and two. T o remove the

ambiguity between logical reasoning levels one and two researchers had to add more conditions

and revise the rubric. If the answer is completely correct with incomplete reasoning it belongs to

level 2, or if complete reasoning but incomplete solution (more than half of the answer) again it

belongs to level 2.

4.6 Guideline and memo for using the analytic rubric over RT1

To ensure that the analytic rubric gives the same output independent of the person who is applying

it, a sample of ten students paper were chosen. The rubric was applied by two different persons

and was adjusted accordingly. In addition, the following scoring guideline was considered to

ensure that the rubric is reliable.

Example one:

57
 a. Since this student has a comprehensive logical argument to support complete correct

solution the logical reasoning score is 3. An example is shown in Figure 16.

Figure 16. Example students work that all categories all complete.

b. This student has a well-organized argument and complete correct notation for a complete

solution, the organization level is 3.

c. This student has a complete use of a graph for producing a complete correct conclusion, so

the use of graph level is 3 as well.

Example two:

a. An incomplete logical argument to support complete correct solution yields logical

reasoning score of 2. An example is shown in Figure 17.

b. An incomplete argument and notation for a completely correct solution yields organization

level of 2.

c. This student did not draw a graph, nor did he/she mention use of a graphing calculator for

making the conclusion, so there is no relevant evidence of the use of a graph for solving

part c. Thus, the level of use of graph is 0.

58
 Figure 17. An example of level two solution.

Example three:

a. A poor reasoning to support an incomplete solution; therefore, the logical reasoning level

is 1. An example is shown in the Figure 18.

b. Some organization for the incomplete conclusion, gives the organization level 1.

c. Use of graph to produce an incomplete solution, so the level of use of graph is 1.

4.7 Memos

Memo 1: Logical reasoning deals with a logical argument to support a conclusion, but organization

deals with an understanding of the concepts (based on w ritten work) as well as correct use of

notations and symbols. For example, for question 3(a) one may find the correct values of zeros

of polynomial but not write the zeros as ordered pairs or x equal forms; therefore, logical reasoning

level would be 3 but organization would be 2.

Memo 2: There are some cases where the scores for all the levels are equal. But it does not mean

that a student who has a poor reasoning level also has a poor use of graph or organization level.

Memo 3: Because of repetition of the following case between student answers for question number

4(d) this memo was made. T he question asked to find the domain. Students usually find

asymptotes and holes, but to show domain, students mostly removed vertical asymptote from real

59
 Figure 18. An example of level 1 solution.

numbers but did not remove hole. In this situation, the logical reasoning level is 2 and it satisfied

complete reason to support an incomplete solution. Organization level is 2 be cause it satisfied

well-organized argument and notation for a correct incomplete solution. An example is shown in

the Figure 19.

4.8 The result of the analytic rubric applied to RT1 and RT2

In the next phase, the analytic rubric was applied to the selected samples. The results of scoring

Figure 19. An example of student work for finding domain

RT1 and RT2 by the analytic rubric were tabulated, which can be seen in Table 18 and 19. SAC

stands for students’ assigned code, LR stands for logical reasoning, or stands for organization

and UG stands for use of a graph.

60
Table 18. RT1 by the analytic rubric.
SAC LR_3a OR_3a Ug_3a LR_3b Or_3b UG_3b LR_3c OR_3c UG_3c LR_4d OR_4d UG_4d
3209 3 3 3 3 3 3 2 2 3 1 1 0
3234 3 2 3 3 3 3 2 2 3 2 2 2
3231 3 2 3 2 2 3 2 2 3 2 2 3
3208 2 2 3 3 3 3 1 1 2 1 1 1
3223 3 3 3 3 3 3 2 2 3 2 2 0
3201 1 1 3 3 3 3 2 2 1 2 2 0
2433 3 3 3 3 3 3 2 2 3 2 2 3
2404 3 3 3 3 3 3 2 2 3 2 2 3
2417 3 2 2 2 2 1 2 2 3 3 2 3
2413 0 1 1 0 0 0 2 2 3 1 1 1
1420 3 3 3 3 3 3 2 2 3 3 2 3
1435 3 3 3 2 2 3 2 2 3 1 2 0
1428 2 2 3 1 1 1 2 2 3 1 1 3
1406 3 3 3 3 3 3 2 2 3 2 2 3
1407 3 3 3 3 3 3 2 2 3 2 2 2
1403 2 2 2 2 2 1 2 2 3 2 2 1
1528 2 2 3 3 3 3 2 2 3 1 2 1
1513 3 3 3 0 1 0 1 1 2 1 2 1
1503 3 3 3 1 1 3 1 1 1 2 2 3
1624 3 3 3 1 2 2 1 1 0 2 2 0
1629 3 2 3 3 3 3 3 2 3 2 2 1
1610 1 1 2 1 1 1 3 2 0 3 3 3
9017 3 2 3 0 0 0 1 1 3 2 2 2
9033 3 3 3 1 1 1 2 1 1 1 1 1
9009 3 2 3 1 1 1 2 2 3 1 1 0

4.9 The result of t-test over mean score obtained using the analytic rubric

As mentioned previously, some common concepts between RT1 and RT2 were selected for a

sample of student papers. The qualitative rubric was applied for these papers for both RT1 and

RT2. In the next phase, the two-sample t-test over the mean scores of the qualitative rubric grades

was applied to find if the use of graphing calculator has a significant effect on students’ logical

reasoning, organization including intellectual order, written order and correct use of symbol and

notation, and use of a graph. A t-test with α= 0.05 and 95 percent confidence interval was used to

compare the following questions from RT1 versus RT2 as well as total logical reasoning, total

61
organization and total use of graph, which includes the mean scores archived by qualitative rubric

of these skills for all questions in RT1 versus RT2.

Table 19. RT2 by the analytic rubric.

SAC LR_1e OR_1e UG_1e LR_3b OR_3b UG_3b LR_3z OR_3z UG_3z LR_3m OR_3m UG_3m
3209 3 2 0 1 1 2 3 2 3 1 1 2
3234 3 3 0 1 1 1 3 2 3 1 1 1
3231 2 2 0 1 1 1 2 2 3 0 0 0
3208 1 1 0 3 2 3 3 3 3 3 3 3
3223 3 3 0 1 1 1 3 2 3 0 0 0
3201 2 2 0 1 1 2 3 2 3 1 1 1
2433 3 3 0 3 3 3 3 2 3 2 2 3
2404 1 1 0 1 1 1 3 2 3 2 2 1
2417 0 1 0 1 0 0 3 2 3 1 1 2
2413 3 3 0 1 1 0 3 2 3 0 0 0
1420 3 3 0 2 2 2 3 2 3 1 1 1
1435 1 1 0 2 2 3 3 3 3 1 1 1
1428 3 3 0 0 0 1 3 3 3 1 1 2
1406 3 3 3 1 2 1 2 1 2 0 0 0
1407 3 3 0 1 0 0 3 3 3 1 1 1
1403 3 3 0 1 1 1 3 1 3 0 0 0
1528 3 3 0 1 1 1 3 2 3 0 0 0
1513 3 3 0 2 2 3 3 2 3 1 1 1
1503 1 1 0 0 0 0 2 1 3 0 0 0
1624 1 1 0 1 0 0 3 3 3 0 0 0
1610 1 0 0 1 1 3 3 2 3 0 0 0
9017 0 0 0 0 0 0 1 1 1 0 0 0
9033 2 2 0 2 1 2 3 3 3 0 0 0
9009 0 0 0 1 1 1 3 2 3 2 1 2

Table 20. t-test over mean scores of the qualitative rubric on review test.
t-test over mean scores of different categories df P-value
of the qualitative rubric
Logical Reasoning 47.828 0.01914
Organization 46.445 0.0001978
Use of graph 45.734 9.113e-08

Overall, there is a significant difference in logical reasoning, organization, and use of a

graph of students’ when they used a graphing calculator (RT1) compared to when the same

students did not use a graphing calculator (RT2). Significance level is α= 0.05 and p-values are

less than .05 for all categories of the qualitative rubric. Therefore, with 95 percent confidence

62
interval the null hypotheses (the mean scores of the qualitative rubric in review RT1 and RT2 are

equal) is rejected and alternative hypotheses (the mean scores of qualitative rubrics in RT1 and

RT2 are not equal) is accepted. In addition, the researcher conducted a two-sample t-test over the

mean score by qualitative rubric for a similar concept in RT1 and RT2. The results of t-test show

that there are significant differences in students’ logical reasoning, organization, and use of graph

skills in question 3b versus 3m, question 3z versus 3b and use of a graph, and organization level

of question 4d versus 1e. But there is not a significant difference between

logical reasoning, organization, and use of a graph for question 3a of RT1 versus 3z of RT2.

4.10 Estimating the population mean and confidence interval of the analytic rubric

The total number of students in the population is N=125. We divided the population into

three strata which contain N1=21, N2=70 and N3=33 and one outlier observations. Samples of n1=4,

n2=14 and n3=7 students are chosen from stratum one, two and three respectively. Let 𝑦𝑦ℎ denote

the mean of each stratum (where h runs from 1 to 3). Therefore, one can write them as:
𝑛𝑛ℎ
1
𝑦𝑦ℎ = � 𝑦𝑦ℎ𝑖𝑖 Equation 1
𝑛𝑛ℎ
𝑖𝑖=1

where 𝑦𝑦ℎ𝑖𝑖 is each observation in the stratum h.

The finite-population variance from each stratum is:

𝑁𝑁ℎ
1
𝜎𝜎ℎ2 = �(𝑦𝑦ℎ𝑖𝑖 − 𝜇𝜇ℎ )2 Equation 2
𝑁𝑁ℎ − 1
𝑖𝑖=1

The sample variance from stratum h can be written as:

𝑛𝑛ℎ
1
𝑠𝑠ℎ2 = �(𝑦𝑦ℎ𝑖𝑖 − 𝑦𝑦ℎ )2 Equation 3
𝑛𝑛ℎ − 1
𝑖𝑖=1

63
With the stratified sampling method, one can find the unbiased estimator of the population mean

as
𝐿𝐿
1
𝑦𝑦𝑠𝑠𝑠𝑠 = � 𝑁𝑁ℎ 𝑦𝑦ℎ Equation 4
𝑁𝑁
ℎ=1

where L =3 is the number of strata.

The variance of the stratified population mean is

𝐿𝐿
𝑁𝑁ℎ 2 𝑁𝑁ℎ − 𝑎𝑎ℎ 𝜎𝜎ℎ2
𝑣𝑣𝑎𝑎𝑐𝑐�𝑦𝑦𝑠𝑠𝑠𝑠� = � � � � � Equation 5
𝑁𝑁 𝑁𝑁ℎ 𝑎𝑎ℎ
ℎ=1

Therefore, the unbiased estimate of the variance of the stratified population mean is
𝐿𝐿
𝑁𝑁ℎ 2 𝑁𝑁ℎ − 𝑎𝑎ℎ 𝑣𝑣ℎ2
𝑣𝑣�
𝑎𝑎𝑐𝑐�𝑦𝑦𝑠𝑠𝑠𝑠� = � � � � � Equation 6
𝑁𝑁 𝑁𝑁ℎ 𝑎𝑎ℎ
ℎ=1

To find the confidence interval one can use the t distribution with a modified degree of freedom.

The 100 (1-α) % confidence interval of the estimator of the mean population will be

𝑦𝑦𝑠𝑠𝑠𝑠 = 𝑦𝑦� 𝑠𝑠𝑠𝑠 ± 𝑐𝑐𝛼𝛼 �𝑣𝑣�

𝑎𝑎𝑐𝑐�𝑦𝑦� 𝑠𝑠𝑠𝑠�
�2 Equation 7

The degrees of freedom of t distribution can be estimated as

(∑𝐿𝐿ℎ=1 𝑎𝑎ℎ 𝑣𝑣ℎ2 )2

𝑎𝑎 =
(𝑎𝑎 𝑣𝑣 2 )2 Equation 8
�∑𝐿𝐿ℎ=1 𝑎𝑎 ℎ−ℎ1 �
ℎ

where 𝑎𝑎ℎ is defined as:

𝑁𝑁ℎ (𝑁𝑁ℎ − 𝑎𝑎ℎ )

𝑎𝑎ℎ = Equation 9
𝑎𝑎ℎ

64
With the mentioned formulation of stratified sampling the estimator of the mean and variance of

the population, and confidence interval of the mean of three features of the students RT1 and RT2

were tabulated and can be seen in Tables 21 and 22 respectively. These three features are logical

reasoning, organization of written work and the use of a graph.

Table 21. RT1.

Estimated Estimated Stratified The confidence
mean standard error degrees of interval of the mean of
freedom the population

Logical Reasoning 8.20 0.34 20 (7.62, 8.79)

Organization of written 8.00 0.28 20 (7.52, 8.48)

work
Use of graph 8.94 0.31 20 (8.40, 9.47)

Table 22. RT2.

Estimated Estimated Stratified The confidence
mean variance degrees of interval of the mean
freedom of the population
Logical Reasoning 6.7972350 0.35 20 (6.19, 7.4)

Organization of written 5.8306452 0.32 20 (5.27, 6.4)

work
Use of graph 5.1932604 0.30 20 (4.68, 5.7)

Interpretation of confidence interval: Based on the results shown in tables 21 and 22 in all cases

of logical reasoning, organization of written work and use of graph the estimated mean of students’

skill is larger in the GC test. This also can be seen in the 95% confidence interval of the mean

which is shown in the last columns of the two tables. For example, the confidence interval of the

mean of the logical reasoning skills is (7.62, 8.79) for GC test and (6.19, 7.4) for the none-GC

test. These two 95% confidence interval do not intersect, which shows that the mean of the logical

reasoning skills of the students increases when they use GC on the test. The same analogy is

65
applicable to the organization of written work and the use of graph skills where confidence

intervals of the mean of the skills scores increase when students use GC.

4.11 A detailed qualitative description and discussion of students’ review tests

Application of the analytic rubric to the review tests shows that students have better logical

reasoning, organization, and use of graph skills when they used a GC compared to not using a GC

for similar problems. Below is a discussion of why and how GC might influence students’ skills.

Some sample papers in RT1 and RT2 will be used in the discussion. Question 3 with multiple

subsets and 4 i n RT1 as well as question 1-part e and 3 with multiple subsets in RT2 will be

analyzed. The following themes will be considered for analyzing students’ written work. These

themes were adapted from the analytic rubric. The researcher used a mixed qualitative-quantitative

method (Creswell, 2002) in which both quantitative data such as test scores and qualitative data

such as interviews and students ‘written work collected, analyzed and reported.

Question 3a, b, and c: Zeroes of polynomial, turning point, and end behavior: Question 3

asked students to find and explain zeroes, turning point and end behavior of a polynomial. Students

were also asked to explain how they produced their answers. The RT1and RT2 were designed with

common concepts on t he two tests to give the possibility to compare students’ performance in

Figure 20.An example of a student solution with using GC

66
presence of technology and without technology. A sample of students’ written work on question 3

in RT1 and RT2 are shown in figures 20 and 21, and the details of the differences between their

solutions with and without technology will be discussed.

Case 1:

1a. Sara’s written work on RT1 (GC) question 3

1b. Sara’s written work on RT2 (Non-GC) question 3

Figure 21. An example of the student solution without using GC

case 1a: The first picture shows Sara’s answer to question 3 of the RT1. Sara used the graphing

calculator to produce a graph for the polynomial. Then she labeled the zeros of the polynomial,

which were the points that the graph meets the x-axis. She labeled the turning point with exact

values on the graph that were the points at which the graph changed direction. Sara described the

end behavior of the polynomial by using the graph, and she showed her understanding by correct

notation. For subparts of question 3 in RT1, she had correct and appropriate use of the graph, and

she was able to make a good connection between these concepts and their graphical representation.

67
She used the correct notation for showing zeroes, she used order pairs to show turning points. For

this question, the logical reasoning level is 3 because she used the complete logical argument to

support a completely correct solution. The organization of her written work is 3 because her written

work is well organized and she has completed correct notation for the correct conclusion, and the

use of graph level is 3 because she has a complete use of a graph for solving all parts of question

3.

Case 1b. This shows Sara’s written work on question 3 in RT2 (no GC). Therefore, the graph of

the polynomial was given in question 3 to minimize the effect of not having a graphing calculator.

Students were asked to find the degree of the polynomial and the sign of leading coefficient which

was an indicator of end behavior of polynomial. In addition, students were asked to find zeroes

and maximum and minimum (like turning points in RT1). Sara mentioned that the polynomial has

degree 3 because the polynomial goes down on one side and goes up on the other side. She was

able to make a connection between the end behavior of the polynomial and the degree. But she did

not have enough reason to support why she picked degree 3. T he logical reasoning level for

question 3(a) is 1 b ecause she has some reasoning to support an incomplete solution; her

organization for this part is 1 because she has incomplete reason to support incomplete solution;

her use of graph level for part a is 2 because she has relevant use of the graph in parts. Although

she had a graph of the polynomial, she was not able to make a connection between the graphical

representation of polynomial and the sign of the leading coefficient. She did not answer part b,

so logical reasoning and organization and use of graph level are 0. In question 3(c) (first part c,

she wrote zeros in the following form: x=0, x=1, x=2, x=4. She probably understood that zeros of

the polynomial are the points that the graph meets the x-axis. The logical reasoning level for this

part is 1 because she has some reasoning to support the correct solution. The organization level is

68
2 because she used the correct notation for a correct incomplete solution, and the level of use of

graph is 2 because she has relevant use of the graph in parts. For the second part, she showed the

max= 1 and min= -3. The polynomial has two local maxima at points (0,0) and approximately

(1.5,1) and two local minima at points (0.5, -5) and (2.5, -2.5) approximately. The organization

level for this part is 1 because she has some reasoning for an incomplete solution. The organization

level for the answer is 1 because she does not have a well-organized argument or complete use of

symbols and notation, and the use of graph level is 1 because she used the graph for an incomplete

solution.

Comparing Sara’s work for Q3 in RT1 and Q3 in RT2: Based on o bservation from Sara’s

written work in RT1 and RT2, her levels of producing an answer for similar concepts with and

without technology are different. In one view she had a better performance when she used a GC.

For example, in question 3 of RT1, Sara’s written work is well organized, neat and coherent. She

has appropriate use of a graph for producing complete correct answers. When she used a graphing

calculator, she answered all parts with apparent confidence (see Figures 22 and 23) and she used

completely correct notations and symbols for all parts. She made a good connection between the

concepts of zeros, turning points and end behavior of a polynomial and their graphical

representation. In contrast, in question 3 of RT2, although she had a graph of the polynomial, she

was not able to derive enough information from the graph to answer different parts of question 3.

She was not able to make a connection between the graphical representation of polynomial and

the degree and the sign of leading coefficient. She did not use correct notation to show local

maxima and minima. It seems that she does not have enough confidence to answer the questions

when she does not have a graphing calculator or when she did not produce the graph of the

polynomial herself.

69
Question 4 and 1: Find the domain of a rational function

Case1

1c. Sara’s paper on RT1 (GC)

Figure 22. Sara’s paper on RT1 with GC

1d. Sara’s paper in RT2 (no GC)

Figure 23. Sara’s paper in RT2 without GC

1c: The first picture shows Sara’s answer to question 4 of RT1. In this question, Sara did not use

a GC to produce answers. Written work for part d as well as parts b, c and e should produce

information enough to find the domain. In part e of question 4, Sara mentioned that she used a

70
graphing calculator to find a vertical asymptote and domain, and she used factoring methods to

find a hole. She found the vertical asymptote correctly, and she used correct notation to show it.

She also found a hole, and she removed vertical asymptote from the domain but not the hole.

Visualizing the rational function could help Sara to remove vertical asymptote from domain, but

the graphing calculator does not show a hole in the graph. If students check the table in GC, then

they can find a hole. The hole is the points in the table that in front of x value for which y value is

undefined. This might be one of the reasons that Sara did not remove the hole because only by

looking at the graph of the rational function, it seems continuous everywhere except at the vertical

asymptote. In this case, the logical reasoning level is 2 because it was complete reasoning to

support incomplete correct solution. The organization level is 2 because it has a well-organized

argument and correct notation for the correct incomplete conclusion, and the use of graph level is

2 because it has relevant use of the graph in parts.

1d. The second picture shows Sara’s written work on question 1 in RT2. In this question, Sara

did not use a GC to produce an answer. Part d of this question will be discussed. In addition, parts

c and f will be considered to see if she was able to use information from these parts to find the

domain. Sara used a factoring method to find the domain. She found vertical asymptote by putting

denominator equal zero. She mentioned that there is no hole probably because she did not see any

common term in numerator and denominator, but she did not remove the vertical asymptote from

the domain. It might be because she did not have the graph that she was not able to see that the

rational function does not continue at vertical asymptote points. In other words, she was not able

to visualize the graph of a rational function to see the function does not have any y value at vertical

asymptotes. The logical reasoning level is 0 because of no relevant argument to support a

conclusion. The organization level is 1 because it has some organization and some correct notation

71
for an incomplete incorrect solution. The use of graph level is 0 be cause there is no r elevant

evidence of the use of a graph.

4.12 Comparing Sara’s work for Q4 in RT1 and Q1 in RT2

Sara’s written work on finding a domain with and without technology is discussed. In the test with

a graphing calculator, she was able to find vertical asymptote and hole, but she only removed the

vertical asymptote from a domain, not a hole. As mentioned before this mistake could be

happening because of the way that GC produces a graph. The TI-84 does not show a hole in the

graph. The only way that one can find the hole by GC, is by looking at the window (table and by

checking the value that is undefined). She was able to make the good connection between the

concept of domain and the graphical representation of it. In contrast, in the question 1 of RT2 in

which Sara did not use a graphing calculator to find the domain, she found vertical asymptote by

putting denominator equal zero and she also found a hole, but she did not remove them from the

domain. She was not able to visualize the domain of the function, and she was not able to make

the connection between the concept of vertical asymptote and domain as well as a hole. Since the

non-GC test was taken after the GC based test, it is surprising that she was able to find the domain

in first test partially correct and in second test completely wrong. Perhaps since she did not have a

clear image of the rational function in the second test it might influence finding the domain for the

function. She was not able to see the image of a rational function on the vertical points which could

cause not removing them from the domain. In the first test, she was not able to see the hole on the

graph because the way that TI-84 produces the graph, and that could be the reason that she did not

remove the hole from the domain. Sara’s performance on the concept of domain was better when

she used GC compared to when she did not use GC. As a tool GC could help students to visualize

72
different mathematical concepts including vertical asymptote, horizontal asymptote, and domain,

which can help students to have a better conceptual understanding of these concepts.

4.13 The result of the analytic rubric applied to the concept test

Data from the concept test were collected as another source for answering the research questions.

Therefore, students’ papers from the concept test were analyzed with the same methods as were

the review tests. As a reminder for the reader concept test was the only written test in college

algebra. It was administered two weeks before the final exam and was designed by the college

algebra coordinator. It constituted 10% of the total student grade in college algebra. Four college

algebra sections were considered as treatment sections where students were allowed to use a GC

on the concept test, and four sections were considered as control because they were not allowed to

use a GC on the concept test. In this phase, the researcher used a stratified sampling method to

select sample size for both the control and treatment groups. Three strata were considered: students

with grades 80 and above, students with grades 60 to 80, and students with grades under 60. Some

questions were common between the concept test and RT1 and RT2. Therefore, the rubric was

applied to questions with concepts common with the review tests to give the researcher an

opportunity to explore students’ understanding based on multiple sources. These questions are 2b,

3c, and 3g.

The result of the analytic rubric applied to control group: The total population of control group

was 131 students with N1=12, N2=55 and N3=64. One-eighth of the population was stratified into

3 strata (as defined above) with 2, 7, and 8 students. Table 23 shows the information about

sampling.

After stratified random sampling, the researcher collected the students’ concept test papers for

control and treatment groups. In the next phase, the analytic rubric was applied to the selected

73
samples.

Table 23. Number of students in each stratum.

Students grade /100 Qualitative grade Population size in non-GC Sample size n=17
/4 test. N=131
80 -100 A, B N1=12 n1=2
60 -79.99 C, D N2=55 n2=7
0-59.99 F N3=64 n3=8

The following memos were used based on the repetition of them between students answers

to help the researcher apply the rubric equally on the concept test for both control and treatment

sample size.

Memo 1: In question 2c, if the upper and lower bound of the interval, and interval notation and

Figure 24. An example of solution with level 3 score by analytic rubric

reasoning to describe domain were correct then LR=3, OR=3, and UG=3.

Memo 2: In question 2, if a student incorrectly showed the interval of [0,225] as [0, 25, 50, 5,

225] then the given LR=1, and OR=1.

Memo 3: If students used parentheses or curly brackets to show domain instead of the bracket, but

the reasoning part was correct then: LR=3 OR=2.

74
 Figure 25. An example of solution scored by analytic rubric

Memo 4: In question 2, which asked about the domain, if the upper and lower bound of the interval

and interval notation were correct but the explanation was not correct then: LR=1, OR= 2.

Memo 5: For question 3c, if a student defines zero as x-intercept without any more explanation

then LR=2, OR=2.

Figure 26. A solution with logical reasoning and organization of written work scores of 2

Memo 6. For question 3c shown in Figure 27, if the reasoning does make sense for describing the

zeros and the notation is appropriate, then LR= 3 and OR= 3. For UG we will check student

explanation in part f to see if the graph was used for answering part c. In this case UG=3.

Memo 7: In question 3g shown in Figure 28, if students have complete correct reasoning with

complete correct use of notation then LR=3 and OR=3. For UG, part f of question 3 will be check.

If a student mentioned use of GC to describe end behavior by making the graph, then UG=3.

75
 Figure 27. Question 3c

Memo 8: For question 3g shown in Figure 29, if a student described the end behavior of the graph

Figure 28. End behavior of a function

by indicating that it falls to the left and the right then LR=1, OR=0.

Figure 29. Question 3g

Memo 9: For question 3g shown in Figure 30, if a student explained the end behavior as x

approaches negative infinity, y approaches to -2, and as x approaches positive infinity, y (instead

r(x)) approaches to -2 then LR=3 but OR= 2 because for OR correct use of notation is required.

76
 Figure 30. Question 3g with LR=3 but OR= 2

The results of scoring the concept test for the control group by the analytic rubric were

tabulated, which can be seen in the Tables 24 and 25. SAC stands for student assigned code, LR

stands for logical reasoning, OR stands for organization and UG stands for use of a graph.

Table 24. The results of applying the rubric.

SAC LR_2b OR_2b UG_2b LR_3c OR_3c UG_3c LR_3g OR_3g UG_3g
1426 3 3 3 3 3 0 3 3 3
2433 2 3 3 3 3 0 1 1 0
2432 2 2 3 1 1 2 1 1 2
2207 3 3 3 1 1 0 1 0 0
1403 3 2 3 0 0 1 1 0 0
1410 3 2 3 3 2 3 2 1 2
2438 3 2 3 1 0 0 1 0 0
2203 2 3 3 1 1 2 1 0 0
9005 3 2 0 0 0 1 1 1 0
2403 1 1 1 2 2 0 1 1 0
2424 2 3 3 0 1 0 1 1 0
2209 1 1 3 1 1 0 2 1 2
9002 3 2 2 3 2 0 0 0 0
9024 1 2 2 0 0 2 0 0 1
1435 3 2 3 0 0 0 0 1 0
1420 3 3 3 3 3 3 3 3 3
1411 1 1 1 2 2 2 1 0 1
1413 0 0 1 2 2 3 1 0 0

77
 4.14 The result of the analytic rubric applied to treatment group

Treatment groups had three strata with population sizes 16, 65, and 88. A fraction of

(approximately) 1/8 of population in each stratum was selected for samples. The sample sizes for

treatment groups are 2, 8, and 7 respectively. Table 25 shows the information about sampling.

Table 25. Sampling information.

Students grade Qualitative grade /4 Population size in GC sample size
/100 test n=17
N=136
80 -100 A, B N1=16 n1=2
60 -79.99 C, D N2=65 n2=8
0-59.99 F N3=55 n3=7

In the next phase, the analytic rubric was applied to the selected samples. The results of

scoring the concept test for treatment group by the analytic rubric were tabulated as seen in the

Table 26 and 27.

Table 26. The results of applying the rubric.

SAC LR_2b OR_2b UG_2b LR_3c OR_3c UG_3c LR_3g OR_3g UG_3g
3131 3 3 3 1 1 3 3 2 3
1529 3 2 3 3 3 3 1 0 0
3204 3 3 3 2 3 3 2 2 2
1616 3 3 3 1 1 3 3 3 3
1633 3 3 3 1 1 3 1 2 0
1538 3 3 3 1 1 2 1 1 0
1527 3 3 3 3 3 3 1 1 1
1638 2 2 2 3 3 3 2 1 2
1605 3 2 3 1 1 3 1 1 0
1631 3 3 3 0 0 0 1 0 0
1615 3 3 3 3 3 3 0 1 0
1602 2 3 3 2 2 3 3 3 3
1528 1 1 2 3 3 2 1 0 0
3235 3 3 3 3 3 3 3 3 3
3123 3 3 3 2 2 3 2 2 2
3107 3 3 3 3 2 3 3 2 3
3206 3 3 3 3 3 3 3 3 3

Result of t-test over mean score of the analytic rubric in the concept test
78
 The rubric was applied to questions with concepts in common with the review tests to give

the researcher an opportunity to explore student understanding. In the next phase, the two-sample

t-test over the mean scores of the qualitative rubric grades was applied with significance level α=

0.05 and 95 percent confidence interval was applied to the concept test for control and treatment

sections. The results are shown in Table 27.

Table 27. t-test over mean scores of the qualitative rubric on concept test.
t-test over mean scores of different categories of the qualitative df P-value
rubric for the concept test
Logical Reasoning 28.25 0.0068

Organization 30.47 0.0046

Use of graph 31 0.00055

The result of the two-sample t-test shows that there are significant differences

in logical reasoning, organization, and use of graph skill of treatment students verses control

sections. Students who used a graphing calculator in the concept test have better logical reasoning,

organization including well organized argument, and correct use of notation and symbol compared

to the students who did not used a GC in the same test.

4.15 Estimating the population mean and confidence interval

The total number of students in the control group is N=131. We divided the population into three

strata that contain N1=12, N2=55 and N3=64 observations. The sample size is n=17 including n1=2,

n2=7 and n3=8 students chosen from stratum one, two and three respectively. With the mentioned

formulation of stratified sampling in the previous part the estimator of the mean and variance of

the population, and confidence interval of the mean of three features of the students in the control

and treatment section in the concept test were calculated and tabulated as seen in Table 28 and 29

respectively.

79
Table 28. Concept test control group.
Estimated Estimated Stratified The confidence
mean standard error degrees of interval of the mean of
freedom the population

Logical reasoning 4.61 0.39 10.10 (3.09,5.33)

Organization of written 4.05 0.38 8..06 (3.45,4.77)

work
Use of graph 4.34 0.49 12.00 (3.46, 5.22)

The total number of students in the treatment sections is N=136. We divided the population into

three strata that contain N1=16, N2=65 and N3=55 observations. The sample size is n=17 including

n1=2, n2=8 and n3=7 students chosen from stratum one, two and three respectively.

Table 29. Concept test treatment group.

Estimated Estimated Stratified The confidence
mean variance degrees of interval of the mean
freedom of the population
Logical reasoning 6.65 0.32 8.76 (6.05, 7.25)
Organization of written 6.36 0.32 8.96 (5.77 ,6.95)
work
Use of graph 7.07 0.45 3.31 (6.04, 8.09)

Interpretation of confidence intervals for the concept test

Based on t he results shown in Tables 28 and 29 in all cases of logical reasoning,

organization of written work and use of graph, the estimated mean of students’ skill in the

treatment sections are higher than students in the control sections. This also can be seen in the 95%

confidence interval of the mean which are shown at the last columns of the two tables. For

example, the confidence interval of the mean of the organization skills in the control section is

(3.45, 4.75) and (5.77, 6.95) for treatment sections. These two 95% confidence interval do not

intersect, which shows that the mean of the organization skills of the students very likely increases

80
(probability more than 95%) when they use GC on the concept test. The same analogy is applicable

to logical reasoning and use of graph skills.

4.16 A detailed qualitative description of students’ concept tests

In this section different students’ answers are qualitatively compared and summarized.

Question 2: Domain

For this question about function and meaning of the domain in terms of the problem, 8 students in

control section answer correctly with complete correct reasoning and complete correct notation to

show the domain. Five students had correct reasoning, but they used parentheses instead of

brackets to show the domain. One student used curly braces to show the domain. One student

showed domain as a set of discrete points and two students had an irrelevant solution. A sample

of work of a student in control section is shown in Figure 31.

Figure 31. A sample of work of student in control section on domain problems.

In treatment sections 13 students used correct logical reasoning and correct use of notation to

explain function and the domain of function. Two students had correct logical reasoning, but they

used parentheses to show the domain, and two students used a set of discrete numbers. A sample

of work of a student in the treatment section is shown in Figure 32.

81
 Figure 32. A sample of work of a student in the treatment section on domain problems.

More students in the treatment sections were able to present well-organized reasoning and correct

use of notation to define the function and to find the domain of the function compared to students

in the control sections (13 vs 7). More students in control sections misused notation to show the

domain of the function compared to the students in the treatment section; five students in control

sections used parentheses while two students in the treatment section did so.

Question 3: Zeros and end behavior of a functions

Zeros: First part of this question asked what the zeros of p(x) tell us about r(x) when

𝑝𝑝(𝑋𝑋)
r(x) = 𝑞𝑞(𝑥𝑥) . In treatment sections 12 students described the zeros as x-intercepts or stated that the

graph goes through those points on the x-axis. One student mentioned that “zeros tell us that the

numerator for r(x) are (x-2) (x+4). These are the x-intercept for r(x). Only 4 students have

irrelevant solutions for this part. In control sections 10 students described zeros as x-intercept, 6

students mixed up zeros with either vertical asymptote or multiplicity or they had irrelevant

solutions. Sample of student answer in treatment sections are shown in Figures 33 and 34.

82
 Figure 33. Sample of students answer on finding zeros in the treatment section.

Figure 34. Sample of students answer in control section

End behavior of the function: Last part of this question shown in Figure 35 asked to describe the

end behavior of function r(x). Only two students in control section described end behavior of the

function correctly with the correct notations and symbols. Four students drew a correct graph

manually based on the given information about zeros and vertical and horizontal asymptotes but

they could not describe the end behavior of the function. They were not able to find logical

connection between the graph of the function and the end behavior of the function. Eight students

in the control sections mentioned as x approaches to positive infinity the function grows and as x

approaches to negative infinity the function decay. Three students conclude the following answer:

“the lading confection is negative, and the highest degree is even therefore the function falls to the

right and to the left.” A sample of students answer to this question in shown in Figure 35.

83
 Figure 35. A sample of student answer in control section

In treatment section 6 students were able to produce correct answer with correct notation. Two

students had correct reasoning buy they used y instead r(x). They mentioned that they used GC to

produce a graph and they used the graph to answer to this question. Four students mentioned the

function falls to the right and to the left and three students mentioned as x goes toward positive

and negative infinity R(x) goes toward positive infinity. A sample of student answer is shown

Figure 36.

Figure 36. A sample of student answer in treatment section.

In summary, students who used a GC in the concept test showed more skills in reasoning,

organization including well-organized argument and correct use of notation and symbol, and use

84
of graph in defending domain, zeros and endeavor of a function compare to the students who did

not use a GC in the same test.

4.17 The result of student attitude survey (SAS)

Results of SAS in college algebra: As a reminder for readers, a survey that was developed by

Tharp (1992) and was used by Merriweather and Tharp (1999) was selected. The chosen survey

had 23 items with the format of typical five-level Likert items with responses of strongly disagree

(SD), disagree (D), neutral or undecided (N/U), agree (A), and strongly agree (SA). One item was

added to the student attitude survey to explore students’ proficiency in using a graphing calculator,

rated from 0 to 10. This survey was completed by college algebra students in the first week and

last week of the semester. After collecting the completed survey, the primary researcher changed

the five-level responses to a range of numerical values from 1 to 5. For example, SD= 1, and SA=

5.

In the first phase of analyzing SAS responses, data from pre-and post-survey in the control

and treatment sections were visualized to have a clear image of students’ responses. The images

are shown in Figures 37 to 38 below.

85
 Figure 37. Control Pre-SAS

Figure 38. Treatment pre-SAS

86
 Figure 39. Control post-SAS

Figure 40. Treatment post-SAS

87
 In the next stage, out of 24 survey items, 13 items were selected based on relevancy to the

research questions. The selected items divided into two groups based on their nature: general items

and personal items. General items derived to three subcategories including the positive view in use

of a GC, positive view about mathematics, and negative view about mathematics. Items from each

category are as the following:

General items
Subcategory 1: Positive view about the use of a GC
Q 4. Graphing calculator makes math fun.
Q 6. Learning algebra is easier if a graphing calculator is used to solve problems.
Q 9. It is important that everyone learn how to use a graphing calculator.

Subcategory 2: Positive view about mathematics

Q 21. Learning mathematics means exploring problems to discover patterns and generalize.

Subcategory 3: Negative view about mathematics

Q 13. Mathematics is boring.
Q 19. Learning mathematics is mostly memorizing a set of facts and rules.

Personal items, which refers to items that have the subject “I”, divided into three
subcategories. The first subcategories included items that have a negative view of the use of a GC.
For example, the GC can be a hindrance and a tool which reduces visualization skill. The second
subgroups are the items that show the level of an individual skill of use of a GC. T he third
subcategory shows students impression of their mathematics abilities.
Personal questions
Subcategory 1: Negative view about the use of a GC
Q 3. The graphing calculator will hinder my ability to understand basic computation.
Q 17. I feel I am cheating myself out of a chance to learn when I use a graphing calculator.
Q 18. If I use a graphing calculator my ability to visualize problems will decrease.
Q 22. I rely on my graphing calculator too much when solving problems.

Category 2: An individual skill of use of a GC

Q 8. I know how to use a graphing calculator very well.
Q24. How much experience of using a calculator in math courses have you had?

Category 3: Personal impression of mathematics ability

Q 12. I am good at mathematics.

88
At this stage of survey data analyzing, means of responses in the above-mentioned categories in

both pre- and post- survey for both control and treatment groups were computed and a two-sample

t-test with a significance level of ∝= 0 .05 and confidence interval of 95 present was applied. In

the second stage, the students’ responses to the general category in the pre-survey and post-survey

for control and treatment sections were compared. The results are given in the tables 30 and 31.

Table 30. T-test of means of responses for pre-survey between control and treatment.
Categories # of questions Mean(C-Pre) Mean(T-Pre) df p-value
Positive view about use of GC 4,6,9 3.68 3.93 162 0.127
Positive view about math 21 3.74 3.75 192 0.93
Negative view about math 13,19 3.06 3.06 195 0.93
Female 52 65

male 34 52

The result of the two-sample t-test over the mean scores of general category questions of pre-

survey does not show any significant differences between students’ positive view in use of a GC,

positive view about mathematics and negative view about mathematics between control and

treatment sections. Both groups have a positive view of the use of GC (the mean is 3.68 in control

sections vs 3.93 out of 5 in the treatment section). Thus, students in both control and treatment

sections believed that use of a GC makes mathematics fun, learning algebra is easier with a GC

and learning how to use a GC is important. In addition, the number of students who believed that

mathematics means exploring problems to discover patterns and generalize is similar in both

groups; the number of students who believed mathematics is boring and is a set of memorizing

rules were equal in control and treatment sections.

A two-sample t-test over mean responses to the same items between control and treatment

students’ post-survey was conducted. The mean responses to items 4, 6, and 9 for the control group

89
is 3.97 vs 3.76 for treatment groups. The results of t-test show significant differences between

students’ positive views in control and treatment groups. The results are shown in the Table 31.

Table 31. Students’ positive views in control and treatment groups

Categories # of questions Mean(C-Post) Mean(T-Post) df p-value
How does using technology 4,6,9 3.97 3.76 159 0.022
affect the organization of
college algebra and calculus
students’ written work?

Positive view about 21 3.7 3.58 159 0.7

mathematics
Negative view about 13,19 3.09 3.05 148 0.74
mathematics
Female 14 46
Male 22 30
Unknown gender 12 7

Although students in both control and treatment groups had a positive view of the use of

GC, students who were not able to use a GC on the concept test had a more positive view about

the use of GC. They believed that having GC would make algebra easier, more fun and the use of

GC is important. Results are in Table 32. There is not any significant difference between control

and treatment sections positive view in post-survey as well as a negative view of mathematics

between the control and treatment sections (subcategory 2, and subcategory 3).

Table 32. T-test over means of the scores, for post-survey between control and treatment.
Categories # of questions Mean(C-Pre) Mean(T-Pre) df p-value

Negative view of use of 3,18,22,17 2.25 2.29 190 0.6639

a GC

Self-impression of the 8,24 3.54 3.62 177 0.5582

use of GC skill
Self-impression of 12 3.36 3.26 184 0.48
mathematics skills
Female 52 65
male 34 52

90
Personal questions: In the next stage, a two-sample t-test was applied to items in the personal

category between results of students’ pre- and post-survey in the control and treatment sections.

The results do not show any significant difference between students’ negative view of the use of a

GC, students in both groups had an average 2.25 out of 5 for the questions that mentioned GC is a

hindrance or GC reduced the visualization skill. The summary of results is shown in Table 33.

Students in both groups have similar skill (at least 3.5 out of 5) of the use of GC. Students in both

control and treatment section had a positive impression of their mathematics skills.

The same analysis was done on the results of the post-survey between control and treatment

sections. The results show that students in the control section had a more positive self-impression

of their mathematics skill. The results are below in Table 33. There is a significant difference (p-

value = .003<.05) between students’ impression about their mathematics skills in the control and

treatment sections. The average of students’ responses in the control sections who believed that “I

am good at math” is 3.67 in post-survey while this average is 3.16 for the treatment sections. The

students’ impression of their skill in using GC in the control group is slightly higher than in

treatment sections. But there are not any significant differences between their negative views on

the use of GC.

Control pre- and post-SAS, general and personal categories, between the control sections

In the next phase, the results of pre-SAS and post-SAS in control sections were compared.

Students’ positive view on the use of a GC increased for students in the control section. Students

in control sections believed that if they would use a GC, then mathematics is more fun, algebra is

easier, and learning how to use a GC is important. Their positive view about the use of GC

increased at the end of the semester. Students’ positive view about mathematics does not change

from the beginning and the end of the semester, but more students in control sections believed

91
“math is boring”, and “learning mathematics is just memorizing the rules” at the end of the

semester. Results are in the Table 34.

Table 33. Students' positive self-impression of their mathematics skill.

Categories Items Mean(C-Post) Mean(T-Post) df p-

value

Negative view of use of GC 3,18,22,17 2.17 2.23 158 0.59

Self-impression of the use of 8,24 3.99 3.74 157 0.76

GC skill
Self-impression of mathematics 12 3.67 3.16 154 0.003
skills
Female 44 46

Male 22 30

Unknown 12 7

The same analysis methods applied to the personal question of the pre- and post-SAS, between

control groups show that students’ impression of their skill in the use of a GC increased at the end

of the semester in control sections. But there is not any significant difference about student’s

impression on their mathematics skill as well as their negative view of the use of GC.

Table 34. Students positive view about mathematics

Categories # of questions Mean(C-Pre) Mean(C-post) df p-value

View about use of GC 4,6,9 3.68 3.97 162 0.003

Positive view about math 21 3.74 3.7 153 0.72

Negative view about math 13,19 3.06 3.09 154 0.003

Negative view of use of GC 3,18,22,17 2.25 2.17 159 0.38

Self-impression of the use of 8,24 3.54 3.99 161 0.002

GC skill
Self-impression of 12 3.36 3.67 161 0.59
mathematics skills

Treatment pre-SAS and post-SAS in general and personal categories

92
 In the next phase, the results of pre- and post-SAS on the general and personal question in

the treatment sections were compared by the same analysis as previous parts. Results are shown

in Table 35.

Table 35. Results of pre- and post-SAS on the general and personal question.
Categories #of questions Mean(T-Pre) Mean(T-post) df p-value
Positive view about 4,6,9 3.93 3.76 161 0.28
use of GC
Positive view about 21 3.75 3.58 158 0.4
math
Negative view about 13,19 3.06 3.05 177 0.94
math
Negative view of use 3,18,22,17 2.29 2.23 170 0.49
of GC
An individual skill 8,24 3.62 3.74 170 0.43
of the use of a GC
Self-impression of 12 3.26 3.16 161 0.59
mathematics skills

The results of t-test over all subcategories of general questions as well as personal questions do

not show any significant difference between treatment students’ view from pre-and post-survey

for the stated categories.

No analysis was attempted based on gender and ethnicity. This could be a topic for future study.

Correlation

In the last stage of SAS data analyzing, students’ level and skill of use of GC and students’

mathematics skill were compared with their positive/negative view about mathematics as well as

their positive/negative views about the use of GC to explore if there is any direct or indirect

correlation between these characteristics. Correlations were calculated by using equation 10.

𝐶𝐶𝐶𝐶𝐶𝐶(𝑥𝑥, 𝑦𝑦) ∑(𝑥𝑥 − 𝑥𝑥̅ )(𝑦𝑦 − 𝑦𝑦�)

𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 = =
𝜎𝜎𝑥𝑥 𝜎𝜎𝑦𝑦 �∑(𝑥𝑥 − 𝑥𝑥̅ )2 �∑(𝑦𝑦 − 𝑦𝑦�)2 Equation 10

where 𝐶𝐶𝐶𝐶𝐶𝐶(𝑥𝑥, 𝑦𝑦) is the covariance of x and y data, 𝜎𝜎𝑥𝑥 and 𝜎𝜎𝑦𝑦 are the standard deviations of x and

y data respectively. Further, 𝑥𝑥̅ and 𝑦𝑦� stand for the mean of data 𝑥𝑥 and mean of 𝑦𝑦 respectively.

93
The number of surveys that were analyzed to check the correlation between the stated categories

was 365 in total including 207 females, 139 males, and 19 students whose gender was not given.

The results are shown in Table 36.

Table 36. Survey Results.

Categories A positive view The negative view A positive view The negative
of mathematics of mathematics of the use of a view of the
GC use of a GC
student’s skill -2% -6.2% 24%
of using GC

Student self- 13.7% -26% 9.4% 8.6%

impression of
mathematics skills

The result show that students’ skills of use of a GC have a positive correlation (24 %) with

students’ positive view of the use of a GC. That means students who have a higher skill in the use

of a GC have a more positive view of the use of GC. Students who were skillful in the use of a GC

believed that GC can make mathematics more fun, make learning algebra easier, and knowing how

to use a GC is important.

By applying equation 10, t he correlation between students’ impression of their

mathematics skills and positive view of mathematics such as “Learning mathematics means

exploring problems to discover patterns” was 13.7 percent. This shows that students who had a

higher positive self-impression of mathematics ability had a more positive view of mathematics.

Students who have a positive impression of their mathematics skills think more positively about

mathematics.

There was a direct correlation between students’ mathematics skills and their positive view

of the use of a GC (positive correlation 9.4 %). This value was calculated by equation 10 and

shows that students who believed “I am good at math” are more likely to believe that GC can make

94
math more fun and makes learning algebra easier. There is a negative correlation (-6.2%) between

an individual skill of use of GC and negative view of mathematics. As students’ skill of use of a

GC increased their negative view of mathematics, such as math is boring, and math is only

memorizing the rules, decreased. Students’ positive view of their mathematics ability is indirectly

dependent on students’ negative view about mathematics with -26% correlation. As students’ self-

estimate of mathematics ability increased the negative view of mathematics such as math is boring,

or mathematics is only memorizing the rules decreased.

4.18 Summary of results of pre- and post-survey in college algebra

Out of 24 s urvey items, 13 items were selected based on relevancy to the research

questions. The selected items divided into two groups based on their nature: general items and

personal items. General items were placed in three subcategories: the positive view of use of a GC,

positive view about mathematics, and negative view about mathematics. Personal items, which

refers to the items that have the subject “I”, divided into three subcategories: a negative view of

the use of a GC, an individual skill of use of a GC, and students’ impressions of their mathematics

abilities. A two-sample t-test with a significance level of ∝= 0 .05 and confidence interval of

95% was applied to the mean scores of students’ responses to the selected items in the pre-survey

between control and treatment section as well as post-survey between control and treatment

sections. In addition, pre- and post-survey in control, and pre- and post-survey in treatment sections

were compared. The results do not show any significant differences between students’ responses

in the control and treatment section in pre-survey for all categories as well as the treatment

students’ responses in pre- and post-survey. The results of t-test over mean scores of general and

personal categories in pre- and post-survey in control sections show that students’ positive view

on the use of a GC increased for students in the control section. Students in control sections

95
believed that if they would use a GC, then mathematics is more fun, algebra is easier and learning

how to use a GC is important. Their positive view about the use of GC increased at the end of the

semester. More students in control sections believed “math is boring”, and ‘learning mathematics

is just memorizing the rules” at the end of the semester. The results show that students’ impression

on their skill at use of a GC increased at the end of the semester in control sections. But there is

not any significant difference on students’ impression of their math skills as well as their negative

view of the use of GC. The same analysis was done on t he results of the post-survey between

control and treatment sections. The results show that students in the control section had a more

positive self-impression of their mathematics skill. The average of students’ answers in the control

sections who believed that “I am good at math” is 3.67 in post-survey while this average is 3.16

for the treatment sections. The students’ impression of their skill at use of GC in the control group

is slightly better than treatment sections. But there are not any significant differences between

their negative views on the use of GC. At the last stage of SAS data analyzing in college algebra,

students’ level and skills of use of GC and students’ mathematics skill were compared with their

positive/negative view about mathematics as well as their positive/negative views about the use of

GC. The results show that students’ skills at use of a GC have a positive correlation (24 %) with

students’ positive view of the use of a GC. The correlation between students’ impression of their

mathematics skills and positive view of mathematics is 13.7%. Students who have a positive

impression of their mathematics skills think more positively toward mathematics.

Students’ mathematics skills are directly dependent on students’ positive view of the use

of a GC (positive correlation 9.4 %). Students who believed “I am good at math” believed that GC

can make mathematics more fun and makes learning algebra easier. Students’ mathematics skills

are directly dependent on students’ positive view of the use of a GC (positive correlation 9.4 %).

96
Students who believed “I am good at math” believed that GC can make mathematics more fun and

makes learning algebra easier. As students’ skill of use of a GC increased their negative view of

mathematics such as math is boring, and math is only memorizing the rules, decreased. Students’

positive views of their mathematics ability is indirectly dependent on students’ negative view about

mathematics with a correlation of -26%. As students’ level of self-estimate of mathematics ability

increased the negative view of mathematics such as math is boring, or mathematics is only

memorizing the rules decreased.

4.19 Results of pre- and post-survey in survey of calculus

The results of pre- and post-survey on the general and personal question between students in

survey of calculus were compared by the same analysis as SAS in college algebra. Twenty-one

students completed the pre-survey -- 13 females and 9 males. Seventeen students completed the

post-survey -- 6 males, 9 females, and 2 unknowns. A t-test over means of students’ responses

in pre- and post-survey was applied. Results are shown in the Table 37.

Students’ positive view about mathematics increased insignificantly at the end of the semester.

The results do not show any significant differences between students view in all subcategories of

general items as well as personal items from beginning and the end of the semester. Since the

sample size is small, other analysis such as finding a correlation between subcategories was

ignored and could be a topic for future research.

4.20 Quantitative data analysis and result in survey of calculus

As a reminder to the readers, in the summer of 2017, two survey of calculus classes with a total of

40 students were chosen and two review tests were administered. The first review test, RT1, which

was designed by the principal researcher, was over the derivative and was taken before the midterm

exam. It has three open-ended questions.

97
Table 37. Students’ responses in pre- and post-survey
Categories # of questions Mean_pre Mean_post df p-value
Positive view about use of GC 4,6,9 3.60 3.63 30 0.92

Positive view about math 21 3.45 3.70 35 0.44

Negative view about math 13,19 3.18 3.00 37 0.48

Negative view of use of a GC 3,18,22,17 2.01 2.01 31 0.98

An individual skill 8,24 3.80 3.79 28 0.98

of use of a GC

Student self-impression of 12 3.17 2.77 34 0.25

mathematics skills

The second review test 2, RT2, which was over continuity, limit, derivative and integrals, was

taken before the final exam and it had 7 open-ended questions. One class was considered as the

control in which students were not allowed to use a GC in RT1 and RT2, and one class was the

treatment section in which students used a GC for both review tests. Sample questions of RT 1 and

RT2 are shown in below.

RT1

Q1: Find the open interval where the function 𝑓𝑓(𝑥𝑥) = −2𝑥𝑥 3 + 12𝑥𝑥 2 + 170𝑥𝑥 − 6 is concave
upward or concave downward. Find any inflection point. (5 points)

Q2: a) Let f (x, y) be a function that has (6, 7) as a critical point. We determine that

𝑓𝑓𝑥𝑥𝑥𝑥 (6,7) = −2 , 𝑓𝑓𝑦𝑦𝑦𝑦 (6,7) = 2 𝑎𝑎𝑎𝑎𝑎𝑎 𝑓𝑓𝑥𝑥𝑥𝑥 (6,7) = −10

What D test tells us about the function f? (5 points)

𝜕𝜕𝜕𝜕
b) Find the partial derivative 𝜕𝜕𝜕𝜕
of 𝑧𝑧 = 8𝑥𝑥 + 7𝑥𝑥𝑦𝑦 3 − 6𝑦𝑦 2

Q3: Find all the local maximum, local minimum, and saddle point of the given function:
𝑓𝑓(𝑥𝑥, 𝑦𝑦) = 4𝑥𝑥 2 + 6𝑥𝑥𝑥𝑥 + 8𝑦𝑦 2 + 4𝑥𝑥 − 20𝑦𝑦

98
RT2

Q2: Find the below limit. Which method did you use (table, graph, or limit properties. Show your work).
(5 points)
𝑥𝑥 2 −36
lim
𝑥𝑥→−6 𝑥𝑥+6

Q3: a) Sketch a graph for the below function. (10 points)

3 𝑥𝑥 < 0
⎧
⎪
(x)= 𝑥𝑥 2 + 1 0 ≤ 𝑥𝑥 ≤ 3
⎨
⎪
⎩ 10 𝑥𝑥 > 3

b) Find all values of x where the function f is discontinuous. (Show all steps of your work)
c) For which x value in the interval [0,3] limit of f(x) exists? Why?

Q6: Evaluate the following integrals. (Show all steps of your work). (10 points)
2 +6𝑥𝑥
a) ∫(𝑥𝑥 + 1)𝑒𝑒 3𝑥𝑥 dx
5 2
b) ∫1 𝑑𝑑𝑑𝑑
(5𝑥𝑥+1)3

Q7: Find the area between the following curves. (Show your work). (5 points)

X=-4, x=3, y=0, and 𝑦𝑦 = 2𝑥𝑥 2 + 4

After grading RT1 and RT2 following the same rubric, in the next step, the principal researcher

conducted a two-sample t-test with a significant level of α= 0.05 and 95 percent confidence interval

over the mean scores of RT1 of the control and the treatment section. The result of the t-test is

shown in the Table 38.

Table 38. Two sample t-test over the mean of the RT1.
Number of questions Mean scores in the Mean scores in the df p-value
control section treatment section
Q1 2.87 3.625 14 .3942

Q2 2.75 4.37 11 .13

Q3 2.5 4.25 10 .0422

99
There is a significant difference between students’ performance on Q3 in the control and treatment

section (p-value = .0422<.05). which means students who used a GC in RT1 had a better result in

finding maximum/minimum or saddle point compared to the students who did not use a GC on the

same question with a 95 percent confidence interval and a significance level of α = 0.05. Although

the mean scores on Q1 and Q2 are higher in treatment sections compared to the control sections,

there are not any significance differences between students’ performance on Q1 and Q2 for both

groups. With the same method, data from RT2 were compared between the control and treatment

section. The result of the t-test over RT2 is shown in the Table 39.

Table 39. The result of the t-test over RT2.

questions Mean scores of control Mean in the treatment section df p-value
sections
Q2 3.6 2.18 19 .17

Q3 1.2 4 17.1 0.02535

Q6 4 4.45 17.3 .7745

Q7 1 2.7 17 .03745

There is a s ignificant difference between students’ performance on Q3 and Q7 of RT2 in the

control and treatment sections. The p-value for questions 3 and 7 is less than 0.05 which indicates

that students who used a GC on RT2 had better results on the concept of limit a nd integral

compared to the students who did not use a GC on the same question with a 95 percent confidence

interval and a significance level of α = 0.05. There are not any significant differences between

students’ mean scores on Q2 and Q6 of both groups.

4.21 The result of the analytic rubric applied to RT1 and RT2

RT1: The results of the t-test over RT1 show that students who used GC on the test had better

achievement on Q3. To explore the effect of the use of a GC on calculus students’ understanding

100
and organization of their written work in Q3 of RT1 between control and treatment sections the

qualitative rubric was applied to Q3. Since the sample size is small, the qualitative rubric was

applied to all student papers for Q3 in both sections. The results are in Table 40.

Table 40. RT1 control and treatment sections by the analytic rubric.
C-LR-3 C-Or-3 T-Ug-3 T-LR-3 T-Or-3 T-Ug-3
1 1 0 3 3 3
2 2 0 3 3 3
1 1 0 2 2 0
0 0 0 3 2 0
1 1 0 3 2 0
1 1 0 3 3 0
3 3 0 1 1 0
3 3 0 3 3 0

RT2

The results of t-test over RT2 show that students in treatment sections had a higher mean score on

Q3 and Q7. To dig deeper on how the use of GC affected students’ understanding and organization

of written work on Q3 and Q7 of RT2, the qualitative rubric was applied to Q3 and Q7 of students’

papers in both control and treatment sections. Since the sample size is small, the qualitative rubric

was applied to all papers for both sections. The results are shown in the Table 41 and Table 42.

Table 41. RT2 control group by the analytic rubric.

LR-3a OR-3a UG-3a LR-3c OR-3c UG-3c LR-7 OR-7 UG-7
1 0 1 0 0 0 0 0 0
1 0 0 0 0 0 1 1 0
1 0 0 0 0 0 0 0 0
1 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
2 2 2 1 1 0 0 0 0
0 0 0 0 0 0 0 0 0
2 2 2 2 1 2 1 1 0
2 2 2 3 3 3 3 3 3
0 0 0 0 0 0 0 0 0
101
The results of the two-sample t-test over the results of the analytic rubric applied to RT1 and RT2

do not show any significant differences between students’ logical reasoning, organization, and use

of graph skills in control and treatment sections.

Table 42. RT2 treatment group by the analytic rubric.

LR-3a OR-3a UG-3a LR-3c OR-3c UG-3c LR-7 OR-7 UG-7
1 1 1 0 0 0 3 3 3
0 0 0 0 0 0 3 3 3
1 0 0 0 0 0 0 0 0
1 0 0 2 2 2 1 1 2
0 0 0 0 0 0 2 2 2
1 0 0 1 0 1 1 0 0
1 0 0 1 0 0 0 0 0
3 3 3 1 1 1 3 3 3
3 3 2 2 2 2 3 3 3
2 2 2 0 0 0 3 2 3
1 0 1 0 0 0 0 0 0

The number of calculus students who took RT1 in the treatment sections was 8 and 10 i n the

control sections. Eleven students in treatment sections took RT2. Since the sample size is small

the t- test would not help to draw any conclusion.

4.22 Discussion of RT1 and RT2 for survey of calculus

Discussion of Q3 of RT1: Find all the local maxima, local minima, and saddle points of the

given function: 𝑓𝑓(𝑥𝑥, 𝑦𝑦) = 4𝑥𝑥 2 + 6𝑥𝑥𝑥𝑥 + 8𝑦𝑦 2 + 4𝑥𝑥 − 20𝑦𝑦.

In the control section four students produced a completely correct solution while five students in

the treatment sections did so. Students in both sections used the second partial derivative test to

solve the problems. One student in control and one in the treatment section used the second partial

derivative test but incorrectly said (-2,2) is a local maximum. The sample of students’ work is

shown in Figure 41 and 42.

102
Although most of the students in both groups were able to answer this question correctly students’

written work in the treatment section was more organized, neater, and shorter than the students in

the control section.

Figure 41. A sample of student papers in the treatment section

Figure 42. A sample of student papers in the control section

103
 4.23 Discussion for Q3 and Q7 of RT2

Q3: a) Sketch a graph for the function below.

3 𝑥𝑥 < 0
⎧
⎪
f(x)= 𝑥𝑥 2 + 1 0 ≤ 𝑥𝑥 ≤ 3
⎨
⎪
⎩ 10 𝑥𝑥 > 3
d) For which x-values in the interval [𝟎𝟎, 𝟑𝟑] do the limit of f(x) exists? Why?

In part (a) of Q3, most of the students in both control and treatment section had difficulty in

drawing the graph correctly. Most of the students drew the graph of y=3, y=𝑥𝑥 2 + 1 , and y=10

for all real numbers not for the stated domains.

Students in both groups had difficulty graphing the piecewise function. They were able to graph

the lines and parabola separately but fail to combine these using the given intervals. A sample

answer is given in Figure 43.

Figure 43. Sample of students answer for piecewise function

In part c, although some of the students were able to find the limit they did not use correct natation

as seen in Figure 44. Students used MLP, which is computer-based testing. They learn by focusing

on the answer, and they do not need correct notation, and they do not need to show their work for

credit.

104
 Figure 44. Sample of students’ answer of limits

Q7: Find the area between the following curves. (Show your work). (5 points)

X=-4, x=3, y=0, and 𝒚𝒚 = 𝟐𝟐𝒙𝒙𝟐𝟐 + 𝟒𝟒

Most of the students in the control section even did not try to answer the question. Only 2 students

out of 10 s tudents in control sections responded to Q7, while almost all students in treatment

sections tried to answer this question. Seven students in treatment sections produced a correct

solution with correct notation. Some of the students in treatment sections drew the graph of

quadratic using their calculator to shade the area under the curve. By using the graph, they were

Figure 45. Sample of students answer in treatment section on definite integral.

105
 Figure 46. Sample of students answer in treatment section on definite integral.

able to find the lower and upper bounds for the integral. A sample of work is shown in Figure 45.

Students in the treatment section used TI-84 to find the answer for the definite integral. Samples

of student work are shown in Figure 46. Giving the final answer to two decimal places that they used GC.

One can infer that the restriction on calculator usage in the control sections significantly affected

Figure 47. Sample of students answer in control section on definite integral.

106
students’ self-confidence or success in evaluating the definite integral. A sample of student work

in the control sections is shown in Figure 47.

In this sample, although the student was able to find the correct limits and the antiderivatives the

calculation is incorrect. One can safely say that using GC will help students in evaluating definite

integrals. A lso, GC can help students find the limits of integrals efficiently, and GC can help

students visualize the curves and have a good understanding of the area above or below the curve.

4.24 Summary of results in the survey of calculus

In summary, there is a significant difference between students’ performance on Q3 in the control

and treatment which means students who used a GC in the RT1 had a better result in finding

maxima/ minima or saddle points compared to the students who did not use a GC on the same

question. The results show a significant difference between students’ performance on the Q3 and

Q7 of RT2 between the control and treatment sections. Students who used a GC on RT2 had a

better result in the concept of limit and integral compared to the students who did not use a GC.

The results of the qualitative rubric applied to RT1 and RT2 do not show any significant

differences between students’ logical reasoning, organization, and use of graph skills in control

and treatment sections. The qualitative analyzing of students’ papers on the stated problems show

that although most of the students in both groups were able to answer Q3 on review test correctly,

students from the treatment section had more organized, neater, and shorter work than the students

in the control section. Question 3 (Q3) of RT2 asked students to graph a piecewise function and

find the limit. Students were able to graph the line, parabola and constant line separately but failed

to combine these using the defined intervals. Although some of the students were able to find the

limit, they did not use correct notation. For example, they used lim = 10 which lacks the function
𝑥𝑥→3

expression after limit sign. One reason for not writing the correct notation could be the use of

107
MLP, which is a computer-based testing. This may lead students to focus on the answer rather than

correct notations. In addition, on MLP students do not need to show their work for credit.

For the definite integral question, only 2 s tudents out of 10 in the control section were able to

produce a correct answer, and the rest of students did not try to answer it. While almost all students

in the treatment sections tried to answer this question and 7 of them produced a completely correct

solution. Some of the students in treatment sections drew the graph of quadratic using their

calculator and shaded the area under the curve. By using the graph, they were able to find the lower

and upper bound for the integral. One can say using GC would help students in evaluating definite

integrals. Thus GC can help students to find the bounds of integrals and calculate the values of

definite integrals. In addition, GC could help students visualize the curves and have a good

understanding of the area above or below the curve.

4.25 Interview

As a reminder to the readers, after all the written tests were taken, some students in both control

and treatment sections of college algebra and calculus were interviewed. Students were asked

general questions about their attitude toward using technology in mathematics courses and specific

questions about the way they used the graphing calculators for solving the problems on written

tests. Five college algebra and three survey of calculus students were interviewed. The audios were

transcribed and coded to summarize students’ attitudes toward GC and their approach to solving

problems. The following codes, important to the research questions, were noted in the transcripts

of college algebra interviews.

• Zeros and y-intercept of a function.

• Domain, vertical and horizontal asymptotes, and holes of a function

• Students’ views of the effect of GC on the organization of written work

108
 • Logarithm

• Were students encouraged to use GC?

• Use of GC in MLP and concept test

Zeros and y-intercept: As the first question was about zeros and y-intercepts, the interviewer

asked about the definition of zeros and y-intercept. Most of the students had a clear understanding

of the definitions of zeros and y-intercept of a function. For example, they mentioned that zeros

are the x-intercepts, i.e. when the graph touches the x axis. The y- intercept is when the graph

touches the y axis. This means that most of the students have a graphical view of zeros and y-

intercept. To find the zeros most of them preferred to use GC and especially using the zero finder

of GC except one of them who calculated manually mentioned that for complicated functions he

uses GC. To find the y-intercept, all of them evaluated the value of the function at x = 0.

Domain: Students defined the domain of a function as the values of x that give appropriate output.

All of them knew that holes and vertical asymptotes should be excluded from the domain. Students

unanimously defined asymptotes as a straight line that the graph cannot touch. Only one added

that the function becomes infinitely close to the line but never touches it. To find the domain of a

function three of them found a hole and asymptotes manually and excluded them from the domain.

Two others used GC to graph the function and used the graph to find the domain. When the

students were asked how to find the domain using GC, one mentioned he would graph it and look

at the graph for holes and asymptotes. Some knew that GC has a limitation in showing the holes

in the function and mentioned that they knew they should be cautious and therefore found it

manually. One mentioned that:

“I just plug it that equation into the y =, and there where shows vertical and horizontal
asymptote. But I think it does not show the hole unless you look at the table function. “

109
The notation was another feature that students were asked about. One mentioned that:

“I actually don’t like the interval notation, so I say x is the elements of real numbers except
it cannot be equal -2 and 03. “

In addition, because of range limitation on t he GC screen some of them mentioned finding

asymptotes using GC is not easy. Nevertheless, most of them checked their work with GC and

even for holes they plugged in the hole value in the function to check their work.

The effect of GC on the organization of written work: All the interviewees had the feeling that

using GC would help them to have more organized written works. This includes those who did not

use GC heavily. However, one of them added:

“I think it will be more organized because you will not work them out all. I feel it will be
more organized but if somebody wants to follow your work maybe hand-written is better
because some steps will be skipped when you use GC”

Students’ encouragement to use GC: All the students mentioned that their instructor encouraged

them to produce answers both manually and using GC.

Using GC in MLP and concept test: Interviewed students mentioned a different issue about using

GC in MLP. One of them mentioned he did not use GC heavily on MLP because MLP software

does not like the format of the numbers from GC, which leads to problems. Another student

mentioned that he tried to use GC on every problem just to check his result. One mentioned that

he used GC more than he does normally and mainly because he wanted to check the results. The

other mentioned that for some concepts such as systems of equations he used GC on the MLP test.

The concept of logarithm: Students were asked about the definition of logarithms. Only two out

of five stated that logarithm is the inverse of an exponential function. One of them knew there is a

connection between logarithms and exponential functions but did not exactly know the relation.

110
She talked about the similarity of converting log of a multiplication to the sum of the logarithms

with converting the multiplication of two exponentials with the same basis to one exponential. The

other two students could not define logarithm and did not know about its relationship with

exponential function. They just mentioned they know the properties and mentioned that the level

of the difficulty of logarithms is like other concepts of this course.

The following codes were noted in the transcripts of survey of calculus interviews.

• Limit

• Derivatives

• Integrals

• General view about GC and Desmos

• Organization of written work

The limit was the first subject of the interview for survey of calculus students. All three of them

mentioned that they did not use GC to answer the limit question. Instead they used the limit

rules. For the question where the graph was given, they thought that GC was not needed;

however, they used the graph to answer the question. Another question about limit was answered

with three different approaches. One answered it using the rules and without a graphing

calculator. Another solved it by GC using a table. He mentioned:

“I put the whole equation in the GC and then used the table and looked around 1 and
made sure there is not any hole near the values. “

The last one graphed the function and calculated the limit manually by rules and checked them

together.

111
Derivatives are the next concept that was discussed. The first question asked students to find the

equation of the tangent line to the graph of a function at a certain point. All three students solved

the problem manually. They found the derivative and evaluated the derivative function to find the

slope. Students were asked to make a connection between extrema of a function and its derivative.

To find the extrema of the functions one graphed the function and traced the graph and visually

located the extremum (unaware of the fact that there may be another extremum out of the screen

boundaries). The other two set the derivative equal zero and found the extrema. Students were

asked to find the slope of the tangent line at a local minimum or maximum and they correctly

mentioned it will be zero. However, when they are asked to talk about the graphical correlation of

the derivative and the original function, one of them did not efficiently say that the value of the

derivative at each point is equal to the slope of the tangent line to the graph. Concavity was also

another feature that was explored. Two of the students used the second derivative to determine

whether the graph of the function is concave down or up. One used a graphical approach where

she graphed the function and looked at the graph to see if it is concave down or up (again unaware

of the fact that there maybe something relevant exists out of the boundaries of the screen). She

stated that:

“I would use my GC and graph the original function and then try to solve it the and then I
do not do the second derivative and just look at the graph to find concave down or up”

Students were also asked about their view of integrals. Two of them could make a connection

between the integral and the area under the graph of a function. When they were asked about the

unknown integral all three of them mentioned it means finding the anti-derivative.

112
One student believed that anti-derivative was the hardest part of the entire course because she must

remember lots of rules, and when she was asked why she did not have a problem with derivative

which has the same number of rules she mentioned that:

“Because my mind was working forward, and it was hard to turn it around!”

Two of the students were asked about their opinion on the effect of GC on the organization of their

written work. Both believed that GC and Desmos had a positive effect on the organization of their

written work. However, one mentioned that sometimes she forgets to write things down, but she

generally likes to write things down, which sometimes introduces some error in the writing

formulas.

In general, all three believe that GC could be helpful in their understanding and they did not heavily

depend on i t. One student in control section also mentioned that integral was harder for him

compared to other topics because they did not use technology that much.

4.26 Summary of response to the four research questions:

In summary, multiple data sources were collected to address the research questions.

Different methods of data analysis and statistical tools were used for analyzing the qualitative and

quantitative data. The findings are evidenced by students written work, student interviews, and

students’ grades on several tests, and students’ responses to attitude survey for college algebra and

survey of calculus courses.

Research question 1. Research question one is about the effect of the use of technology on

students’ understanding and achievements in college algebra and survey of calculus courses. The

results of the concept test, RT1 and RT2 indicate that college algebra students have a greater mean

score when they used a graphing calculator compared to the time that they did not use a graphing

113
calculator on the similar mathematical concept. Students who used a GC in the concept test had a

higher mean score compared to students who did not use a GC in the same test. In addition, survey

of calculus students who used a graphing calculator had a higher mean score in RT1 and RT2

compared to those who did not use a graphing calculator on the same tests.

Research question 2. Research question 2 asks “what areas of college algebra and calculus are

affected more by technology?” There are several areas that were affected significantly by using

technology. College algebra students who used graphing calculators have a better understanding

of x-intercepts and y-intercept, domain of a function, end behavior, vertical and horizontal

asymptote. However, the performance of students on function composition was similar.

Survey of calculus students who used a graphing calculator have a better understanding of finding

maximum and minimum for two variable functions. They also have a better understanding of the

concept of limit and definite integrals. However, students had similar performances on derivative

problems, indefinite integrals and limits that need the use of rules.

Research question 3. Question 3 asks how using technology affects the organization of college

algebra and survey of calculus students’ written work. The designed qualitative rubric has three

aspects of students written work which is i) reasoning ii) written order iii) use of symbol and

notation. The results of qualitative rubric applied to RT1 and RT2 and the concept test show that

there is a significant difference between logical reasoning, written order, and correct use of symbol

and notation of students when they used graphing calculator compared to the time that they did

not use. Students’ written work is more organized, neater, and they use more correct notation when

they used a GC in their test. They also able to derive more information from a graph that they

produced by a GC themselves compared to the time that the graph was given in the test. However,

no significance difference was observed for the case of students in the survey of calculus class. In

114
addition, all the interviewed students believe that use of technology enhances the organization of

their written work.

Research question 4. Does the use of technology positively impact college algebra and calculus

students’ attitudes toward their mathematics skills? The results of pre- and post- surveys of the

control sections show that students’ negative view of the use of GC, self-impression of the use of

GC skills, and their positive view of mathematics did not change significantly. Nevertheless, their

positive view of the use of GC increased. In the treatment section no significance difference was

observed between pre- and post-survey in all mentioned features. No significance difference

between students view to all sub-categories of pre- and post-survey was observed.

The results reveal that there is a p ositive correlation between students’ self-impression of

mathematics skills and their positive view of the use of a GC.

115
Chapter 5 Logarithm

5.1 Introduction

Mobile learning (m-learning) is the use of mobile or wireless devices such as smartphones,

tablets, and laptops for learning and teaching (Traxler & Kukulska-Hulme, 2005). With the mobile

revolution in the recent years, researchers in different academic areas have considered the potential

use of mobile devices in education. However, the use of these devices in educational research has

not been as rapid as technological development. In fact, the development of apps in everyday life

has been so fast that apps are widely used by students in their education, and often teachers are

behind the students in using apps. Nevertheless, research on the effectiveness of apps in education

has gained momentum in recent years (Park, 2011; Handal et al., 2013; Hargis et al., 2014).

5.2 Purpose of the study

From our own teaching experience and through discussions with other instructors, we

found that most college algebra students struggle with the topic of logarithms. Therefore, this study

was designed to figure out a more effective way to teach the concept of logarithms. The purpose

of this research is to investigate the effect of using tablet and smartphone apps on student learning

when teaching the concept of logarithms to college algebra students.

5.3 Research questions and hypotheses

The research questions for our study are:

1. Does the use of mobile apps in teaching influence college algebra students’ learning

achievement?

2. In what areas of students’ understanding of logarithms is the use of mobile apps most

and least effective?

116
The null hypothesis (H0) states that the mean score of college algebra students on a written test is

unchanged by using mobile apps. (𝜇𝜇𝜇𝜇 = 𝜇𝜇𝜇𝜇). Alternatively (Ha) states that the mean score of

college algebra students in a written test is greater when they did use mobile apps (experiment)

compare to the time that they did not use this technology (control) in a written test 𝜇𝜇𝜇𝜇 > 𝜇𝜇𝜇𝜇.

5.4 Literature Review

The use of technology in education traces back to many studies on the effect of graphing

calculator on students’ learning. Mobile devices have many features that potentially can be used

to enhance learning. Visualization of the concepts, providing instant feedback, student

engagement, and self-directed learning are some of the features of mobile learning (Sung, Change,

& Liu, 2016). An emerging number of research studies have explored the effectiveness of mobile

devices and specifically smartphone apps in mathematics education (Blair, 2013; Handal, El-

Khoury, Campbell, & Cavanagh, 2013b; Hargis et al., 2014). Smartphone or tablet apps have more

potential compared to fixed computers in lab because they are more aligned with the preferred

method of students as they have physical touch, trial and error and ease of use. Many studies have

been conducted on the quality of apps in mathematics education. However most of these studies

are general review of numerous apps (Highfield & Goodwin, 2013; Larkin, 2013) or suggestion

on the methods and criteria to choose apps for classroom (Cherner, Dix, & Lee, 2014; Handal, El-

Khoury, Campbell, & Cavanagh, 2013c; Park, 2011). For example, Handal et al. (2013)

categorized mobile apps in mathematics education based on f unctionality. These efforts help

teachers in selecting effective apps from the many that are available. Most of the studies on the

effect of apps on teaching math are in K-12 grades. The research on apps in college mathematics

education is much limited compared to the use of apps in other disciplines such as engineering and

117
language. T herefore, in this study we focus on the use of apps in teaching college level

mathematics.

Based on our teaching experience and as also reported in the literature (Kenney, 2005;

Larkin, 2013), we found that college algebra students have difficulty in learning logarithms. To

address this issue some research studies focused on students’ view of logarithm. Weber (Weber,

2002) introduced constructions that help students understand the concept of logarithm. This study

concluded that most of students cannot think of logarithm as a process. Berezovski and Zazkis

(2006) argued that most students do not treat a logarithm as a number, and they try to simplify

terms like log 2 3. They may try to reduce it to a fraction. These authors suggest that logarithms

such as log 2 3 should be treated as an object. They also suggested using the exponent definition

for logarithms and developing the logarithmic rules based on exponents. Kenney and Kastberg

(2013) interviewed students about their knowledge of logarithms and found that those who

memorize logarithm rules are more likely to make mistakes in reconstructing or applying the rules.

After reviewing the literature, one can see that logarithms have many features that are confusing

for students. Logarithm notation is different from other concepts and because it is the inverse of

exponential function, the concept of inverse is also a source of confusion for many students

(Weber, 2002). Therefore, the dual nature of logarithm function, as a process and an object, makes

the understanding of logarithm difficult for new learners (Kinzel, 1999; Sajka, 2003). In this study,

we have investigated the effect of teaching logarithms with mobile apps on students’ understanding

of this concept.

5.5 Method

Participants and Setting: One hundred forty-three undergraduate students enrolled in four

different sections of the college algebra course participated in this study. These course sections

118
were randomly selected and consent from the students to participate in this study was then

obtained. Participants were freshmen with various majors from a diverse student body. Two

sections of the course were identified as treatment groups and were asked to use educational apps

during class activities when teaching the topic of logarithms. The two remaining sections served

as control groups. The study took place in the fall 2016 at the University of Arkansas.

Design of the study: Students received five weeks (total of 15 hr s.) instruction in logarithmic

functions as part of their normal college algebra topics sequence during this study. Concepts that

were covered during this five-week period included: exponential and logarithmic functions,

logarithms and their properties, graphs of exponential and log functions, solving logarithmic and

exponential equations. These concepts were introduced to students by lecture and PowerPoint

slides for both treatment and control groups, which started with basic concepts such as definition,

graphs, and properties. Two instructors taught these four sections. Each instructor taught one

control section and one treatment section. Both treatment and control sections were given the same

test for assessment. Most of the examples presented during the lecture were the same for all

sections except when the instructors used websites (e.g. Khan Academy).

Control sections: In the two control sections traditional lecture method of teaching was used.

Power-point slides were used to introduce definitions and show examples. Worksheets were given

to students for independent or group practice. Instructors introduced logarithmic concepts

including definitions, relation to exponential functions, domain, and basic properties. Product

rules, quotient rules, and power rules were taught. Instructors worked out examples on the board

to show students how to solve logarithmic and exponential equations. The growth and decay

applications of logarithms to real life problems were also presented. In addition, expanding and

119
simplifying logarithmic expressions were covered. At the end, students were assessed and

interviewed.

Treatment sections: In the two treatment sections the instructors introduced smartphone

applications as part of teaching and used them during classroom activities. The students, after

having worked on exercises that existed in the apps, were asked to send the screenshots of their

results to their instructors. Details of the applications used in the treatment sections are included

below.

The Logtrainer application is a tutorial and practice-based app that has multiple-choice

questions. In this app, questions were mostly about converting logarithms to exponentials. After

clicking on one of the answer choices for a question, the students could see the correct answer and

a complete explanation of a similar problem. In this app, similar problems were repeated several

times to help students understand logarithms by using their knowledge of exponentials. Sample

problem from this app are shown in Figure 49.

Figure 48. A Sample problem on apps.

120
 In other sessions, students used the Logarithms app’s activities. This app consists of four

parts: logarithm rules, simplification, expansion, and solving logarithmic equations. Each part

includes some multiple-choice problems relating to the topic. Students were able to see correct

answers immediately after picking one of the answer options. The complete solution to each

question was provided in this app; therefore, students were able to review their work and correct

themselves. Sample problems from this app are shown in Figure 50.

Figure 49. Sample problem from Logarithm app.

5.6 Data collection

Data from multiple sources were collected to address the research questions. Some of these sources

were standardized/placement tests that were taken before students began attending college and

some of them were designed by the course coordinators as part of the material for the college

algebra course. The following are all the data sources of this study:

121
Data from ACT scores: ACT scores of college algebra students were collected for both control

and treatment sections.

Data from written test: Students’ written exam that assessed the topic of logarithmic functions

was collected for both control and treatment sections. Only 59 students in treatment and 49 in the

control group section took the test. The written test was designed by two instructors and the portion

of the test which assessed the topic of logarithms consisted of three questions shown in Figure 3.

The questions were designed in view of Bloom’s definition of six levels of intellectual behaviors

(Krathwohl, 1956). The questions were valued at 10, 5, a nd 5 poi nts, respectively. All three

questions were open response and students were not allowed to use a graphing calculator, but they

could use scientific calculators. This test was graded based on the same rubric for all sections.

Q1 What is a logarithm? (10 points)

Q2 Produce an argument that could convince a friend of the following. (5 points)
log 27 27
a) log 𝑏𝑏 ( 𝑀𝑀 + 𝑁𝑁) ≠ log 𝑏𝑏 𝑀𝑀 + log 𝑏𝑏 𝑁𝑁 b) log2 ≠ log 2 ( 9 )
29

Q3 Write an expression as a sum and/or difference of logarithms. (5 points)

𝑥𝑥 3
𝑎𝑎) log 4 ( 𝑧𝑧. 𝑦𝑦) b) log 5 ( 2𝑥𝑥8𝑦𝑦 3 )
√
Figure 3. Test Questions

Data from student interviews: Ten students in control sections and ten from treatment sections

of college algebra were interviewed separately. These semi-structured interview sessions were

20-minutes long, conducted face-to-face, and were audio-taped with students’ permission.

The purpose of interviews was to explore the level of students’ understanding of the

concept logarithms and exponential functions. In addition, the interviews examined students’

attitudes toward using technology, specifically, apps in learning logarithms. General questions

about using technology in mathematics courses were asked. In addition, students were asked

specific questions about the way that they used technology for solving a logarithm problem and

122
problems in the written test. Some questions were asked to clarify how students understood

concepts such as functions and logarithms. Student interviews were transcribed and analyzed

qualitatively using open coding (Corbin & Strauss, 2008).

5.7 Data analysis

To answer the first research question (Does the use of mobile apps in teaching influence

college algebra students’ learning achievement?), a statistical two-sample t-test with significance

level of α=0.05 was performed on mean of students written test scores of the treatment sections

versus the control sections. In addition, the same t-test was performed on the ACT scores of the

treatment and control sections to check if the control sections and the treatment sections have the

same level of pre-knowledge.

To answer the second research question (In what areas of students’ understanding of

logarithms is the use of mobile apps most and least effective?), students’ written exams were

qualitatively analyzed using open coding (Corbin & Strauss, 2008). For this purpose, we collected

different types of students’ answers and coded them. Codes were then combined and categorized

into five groups based on frequency of answers. Moreover, the mistakes were categorized in the

same way into two common types. Interview data were also used to answer research question 2.

Students’ interview transcripts were also coded from which three main categories emerged i)

students’ understanding of logarithmic functions, ii) effect of technology on learning logarithms,

and iii) students’ understanding of logarithmic properties.

5.8 Results and discussion

Result from written test and ACT scores: The mean of the scores on the written test for the

control section was 12.42 while for the treatment section was 14.16 which shows improvement in

the scores of the treatment section. These results show that there is a significant difference between

123
the mean score of treatment and control sections, which implies that students who used apps in

class activities performed better on the written test. Most improvement was observed on the first

question. However, there is not enough statistical evidence to show significant difference in the

performance of students on the other two test questions. The summary of t-tests is shown in Table

43.

Table 43. Mean and p-value for the control and treatment sections
Number of Mean of Frequency of Mean of Frequency Degrees P-
Questions: the control the control the of the of value
sections sections treatment treatment freedom
sections sections
Q1 4.82 49 6.31 59 103.92 0.0109
Q2 3.42 49 3.64 59 101 0.5117

Q3 4.17 49 4.20 59 101.96 0.8693

The means of ACT math scores were also compared, and we found no significant

difference between the pre-knowledge of students for both groups.

5.9 Qualitative findings

Types of students’ answers: On the first exam question (Q1) students were asked to describe a

logarithm. All students’ written tests were coded for both control and treatment sections. Students’

answers for Q1 were coded as one of the five types: i) logarithm as a function, ii) logarithm as an

inverse of an exponential function, iii) logarithm as an exponent, iv) logarithm as an equation and

v) logarithm as properties and rules. The frequency of students’ answers into each category are

shown in Table 44.

The number of students who described logarithms as an inverse of the exponential function

were significantly different between the control and treatment sections. More students in the

control sections described logarithm as an equation.

124
Table 44. Classification of students’ correct responses for Q1.
Q 1: What is a logarithm? Number of students from the Number of students from
treatment sections (N=59) the control sections
(N=49)
Defined logarithm as a function 20 19
Described logarithm as an inverse 33 21
of an exponential function
Described logarithm as an 13 10
exponent
Described logarithm as an 4 7
equation
Used logarithm rules to describe 7 7
logarithms

Students who used apps in class activity were able to make a better connection between

logarithms and exponential functions while students who were taught using worksheets and

traditional lecture method described logarithms as an equation.

Students’ answers on the second test question were coded and classified as one of two

categories – counter-examples and logarithmic rules. Answers that used a counter-example to

verify each statement were classified under category 1, and answers that depended on logarithm

rules (product and quotient rules) were classified under category 2. Part a of the 2nd test question

asked students to present an argument to verify log 𝑏𝑏 ( 𝑀𝑀 + 𝑁𝑁) ≠ log 𝑏𝑏 𝑀𝑀 + log 𝑏𝑏 𝑁𝑁. A few students

tried to compare log 𝑏𝑏 𝑀𝑀 + log 𝑏𝑏 𝑁𝑁 with the product rule and they came up w ith log 𝑏𝑏 ( 𝑀𝑀𝑀𝑀);

therefore, they explained that the left hand side of the statement does not represent the product rule

and therefore the statement, log 𝑏𝑏 ( 𝑀𝑀 + 𝑁𝑁) ≠ log 𝑏𝑏 𝑀𝑀 + log 𝑏𝑏 𝑁𝑁, is correct. Four students in each of

the control and treatment sections used a counter-example to verify the statements. The number of

students who used logarithm rules were similar for the control and treatment sections, 36 versus

33. Frequency of the solution types is shown in the Table 45.

125
Table 45. Categories of students’ responses for Q2.
Q 2:𝑎𝑎) log 𝑏𝑏 ( 𝑀𝑀 + 𝑁𝑁) ≠ log 𝑏𝑏 𝑀𝑀 + log 𝑏𝑏 𝑁𝑁 Number of Number of
log 27 27
students from the students from the
b) log2 ≠ log 2 ( 9 ) treatment sections control sections
29
(N=59) (N=49)

Answer that used a counter-example to verify the 4 4

statement.
Answer that used logarithm rules to verify 36 33
statement.

In question 3, s tudents were asked to write an expression as a sum and/or difference of

logarithms. Students commonly misapplied the power rule. Table 46 gives a frequency of students’

responses.

Table 46. Students’ most common mistake

Q3: Most common mistake:
log 5 ( 2𝑥𝑥8𝑦𝑦 3 ) log 5 ( 2𝑥𝑥8𝑦𝑦 3 )= log 5 ( 2𝑥𝑥)+ 3log 5 ( 8𝑦𝑦)

Number of students in the treatment sections 20 students

who made a mistake applying the power rule.
Number of students in the control sections 30 students
who made a mistake applying the power rule.

Students either misapplied the power rule or they did not simplify the logarithm. In the

control sections 30 out of 49 (61%) students misused the power rule, while only 20 out of 59 (34%)

of students in the treatment sections misapplied the power rule.

5.10 Interview findings

A sample of ten volunteer students (5 from control and 5 from treatment) were interviewed about

the three questions on t he test. Students were asked to explain their understanding about the

concept of logarithms. Five students in the control sections mentioned that they just do not really

126
know what logarithm means, they only know logarithm properties, and they can apply logarithm

rules to solve problems. For example, one student mentioned

“I don’t know what it is, but I know the logarithm properties.”

Four students from the treatment sections defined logarithm based on exponential functions, and

they were aware of the relationship between logarithms and exponentials. For example, one

mentioned that:

“A logarithm is the inverse operation to an exponential function. It represents a power to

which the base is raising”.
One student used the graph of logarithm to define it. Students in the treatment sections were asked

if using apps helped them to have a better understanding of logarithms and logarithm rules. All

interviewed students mentioned apps helped them to learn better but one of them mentioned that:

“I think I used them and they were useful for me but when you are on your phone and doing
math you will get distracted”

When students in the control and treatment sections were asked to explain why

log 𝑏𝑏 ( 𝑀𝑀 + 𝑁𝑁) 𝑎𝑎𝑎𝑎𝑎𝑎 log 𝑏𝑏 𝑀𝑀 + log 𝑏𝑏 𝑁𝑁 are not equal, all student responses referred to the product

rule. They mentioned that they only memorized the rules and do not know reasons behind them.

Students were asked about the definition of logarithms. Only two out of five stated that logarithm

is the inverse of an exponential function. One of them knew there is a connection between

logarithms and exponential functions but did not exactly know the relation. She talked about the

similarity of converting log of a multiplication to the sum of the logarithms with converting the

multiplication of two exponentials with the same basis to one exponential.

5.11 Discussion and Conclusion

To investigate the effect of using tablet and smartphone apps on student learning when teaching

the concept of logarithms, we analyzed written tests, data from ACT scores, and interviews that

127
were collected from students enrolled in four different college algebra sections. Our aim was to

figure out a more effective way to teach the concept of logarithms in introductory college math

courses. The results of this study reveal that smartphone applications can facilitate learning of

mathematical concepts as they helped students’ learning of logarithms and exponential functions.

During, this study, researchers were aware of students’ difficulties in understanding the

concept of logarithms. The results of the study indicate that teaching logarithms using apps will

have positive effects on s tudents’ understanding of logarithms. One of the positive effects is

making a better connection between logarithms and exponentials. In this study 56% of the students

in experimental sections (using apps) described logarithms as an inverse of an exponential

function, while the percentage for the control sections were 43%. Furthermore, there was a

significant difference between students’ performance in the control and experimental sections in

applying the power rule. The power rule was misapplied by 34% of students in the treatment

sections, and 61.2% by students in the control sections. In addition, the findings of the interview

data indicate that students felt that the using apps helped them better understand the concept of

logarithms.

Smartphone applications are a relatively new technology that can potentially help students in

understanding challenging mathematical concepts. Despite the use of smartphone apps in K-12

mathematics and college level in non-mathematics courses, the use of this technology in college

level mathematics is very limited. Our study shows that smartphone applications can enhance

students’ understanding of logarithms, which has been reported and observed as a challenging

concept. The dual nature of logarithms as an object as well as a process makes the understanding

of this concept even more problematic for students in introductory math courses (Kinzel, 1999;

Sajka, 2003). Our study provides a starting point for considering the use of apps in college-level

128
math courses such as college algebra to help students develop a better understanding of

mathematical concepts such as functions and logarithms.

129
Chapter 6 Summary, conclusions, and recommendations

6.1 Summary

This mixed qualitative and quantitative methods study addressed the effect of technology

on college algebra and survey of calculus students’ understanding. This research study was

conducted in fall 2016 on eight college algebra classes with a total of 315 students, and in summer

2017, in two surveys of calculus classes with a total of 40 students at the University of Arkansas.

College algebra students who were mostly freshman, participated in three, fifty-minute class

meetings every week. Survey of calculus was taken by business students or students who do not

plan to take further calculus courses. Both college algebra and survey of calculus were three-

semester credit hour courses. Four college algebra classes were considered as control sections in

which students did not use a graphing calculator (GC) on a concept test and four sections as

treatment in which students used a GC on the same concept test. All college algebra sections used

a GC on one review test, RT1, and did not use a GC on a second review test, RT2.

Several sources were used to collect data. A pre- and post- student attitude was

administrated during the first and last week of the semester for both college algebra and survey of

calculus courses. Students’ scores and paper work on three written tests (RT1 and RT2 and concept

test) in college algebra and students’ scores and paper work on two written tests (RT1 and RT2)

in survey of calculus were collected. The concept test was the only paper test normally

administered in college algebra. There were seven open-ended questions designed by the

coordinator of the course. Both RT1 and RT2 were designed to give the principal researcher more

data on students’ written work. The RT1 was a graphing calculator-based test and RT2 was a non-

graphing calculator-based test including some open-ended problems. Some common concepts

were included on RT1 and RT2.

130
The RT1 was taken just before the midterm exam and the RT2 was taken just before the final

exam. A week before the final exam, after all the written tests were taken, a few students were

interviewed in both college algebra and survey of calculus.

A pre- and post- student attitude survey had 24 items was completed during the first and last

week of the semester for both college algebra and survey of calculus courses. Thirteen of the 24

survey items were selected for analysis based on relevancy to the research questions. These 13

were divided into general items and personal items. General items were devided into three

subcategories: positive view in use of a GC, positive view about mathematics, and negative view

about mathematics. Personal items, items that have the subject “I”, divided into three subcategories

including a negative view of the use of a GC, an individual skill of use of a GC, and students’

importation of their mathematics ability. The results of the mentioned categories were compared

in several ways such as control vs treatment sections.

Quantitative and qualitative data analysis enables the researcher to discuss on the effect of

technology on students’ understanding and organization of their work. This research study was

guided by the following research questions.

1. How does the use of technology affect college algebra and calculus students’ understanding

and performance?

2. What areas of college algebra and calculus are affected more by technology?

3. How does using technology affect the organization of college algebra and calculus students’

written work?

4. Does the use of technology positively impact college algebra and calculus students’

attitudes toward their mathematics skills?

To answer research question one, two sources of data were used. First, qualitative data

131
from students’ interviews that were transcribed and coded, based on open coding methods. The

following codes were noted in the transcripts of college algebra interviews: zeros and y-intercept

of a function, domain, vertical and horizontal asymptotes, and holes of a function, students’ views

about the effect of GC on the organization of written work, logarithm, how many students were

encouraged to use GC, use of GC in MLP and concept test. For survey of calculus the following

codes were noted: students understanding of limits, derivatives, and integrals; students’ general

views about the use of GC and Desmos, and students’ views on the effect of GC on the organization

of their written work. In addition, data from the analytic rubric that was applied to students’ written

work was used. The designed qualitative rubric has three aspects of students’ written work, which

are i) reasoning ii) written order iii) use of symbol and notation. This analytic rubric was used to

investigate whether the use of GC influences students’ performances on the mentioned skills. To

apply the analytic rubric on students’ papers proportional stratified random sampling was used to

select the sample size. A memo on how to use an analytic rubric to score students’ papers was

provided and a statistical two-sample t-test was conducted over mean score obtained using the

analytic rubric between control and treatment sections. To answer the second part of research

question one, the effect of technology on students’ performances from students’ grades on RT1,

RT2 and the concept test between control and treatment sections were compered by statistical two-

sample t-test.

In answering research questions two and three, multiple sources of data were used. For example,

the result of data from students’ grades in RT1 and RT2, data from analytic rubric applied to

students’ written work and data from students’ interviews.

132
The fourth research question, which asked about the effect of technology on students’ impression

of their mathematics skills, was answered by the result of data from students pre- and post-survey

as well as students’ interview.

Results from research question one showed that the college algebra students had a greater mean

score when they used a GC compared to the time that they did not use a GC on the similar

mathematical concepts. Students who used a GC on the concept test had a higher mean score

compared to students who did not use a GC in the same test. In addition, survey of calculus

students who used a GC had a higher mean score on RT1 and RT2 than those who did not use a

GC on the same tests.

The results applied to research question two reveals that college algebra students who used GCs

have better understanding of x- and y-intercept, domain of a function, end behavior, vertical and

horizontal asymptote. However, the performances of students on function composition and word

problems were similar. In addition, college algebra students who used smartphone applications in

their class activities were able to make better connections between logarithms and exponential

functions. They also were able to use the power rule more accurately compared to students who

did not use apps.

Survey of calculus students who used a GC have a better understanding of finding maximum and

minimum for functions of two variables. They also have a better understanding of the concepts of

limit and definite integrals. However, students had similar performances on derivative problems,

indefinite integrals and limits that need the use of rules.

The results applied to research question three, based on the results obtaining by applying analytic

rubric to RT1 and RT2, and the concepts test, show that there is a significant difference between

logical reasoning, written order, and correct use of symbol and notation of students when they used

133
GC compared to the time that they did not use a GC on the similar tests. Students’ written work is

more organized, neater, with more correct notation when they used a GC on their test. They also

able to derive more information from a graph that they produced by a GC themselves compared to

when the graph was given in the test. However, no significance difference was observed for the

case of students in the survey of calculus class. In addition, all the interviewed students in college

algebra and survey of calculus courses believed that use of technology enhances the organization

of their written work. Even students who mentioned that they preferred to work on pr oblems

manually still believed that use of a GC would positively affect the organization of their written

work.

The answer to research question four that asked about the effect of technology on students’

impression of their mathematics skills shows that students’ self-impression of the use of GC skills,

negative view of the use of GC, and positive view about their mathematics skills did not change

significantly from pre-survey to post-survey in the control sections. Nevertheless, their positive

view of the use of GC increased. In the treatment sections no significance difference was observed

between pre- and post-survey in all mentioned features. No significant difference between

students’ views to all sub-categories of pre- and post-survey were observed as well.

The results of analyzing students’ responses to surveys reveal that there was a positive correlation

between students’ self-impression of mathematics skills and their positive view of the use of a GC.

Students who believed “I am good in math” believed GC can make math more fun and makes

learning algebra easer. Students’ mathematics skills appear directly correlated to students’ positive

view of the use of GC. In addition, students’ skills on use of a GC have a positive correlation with

students’ positive view of the use of GC. As students’ skill of use of a GC increased their negative

view of mathematics, such as math is boring, and math is only memorizing the rules, decreased.

134
Students who had higher skill in the use of a GC believed that GC can make mathematics more

fun and makes algebra easier.

In summary, results from the study exposed evidence that used of technology (GC, Desmos, and

apps) in teaching and learning increased college algebra students’ understanding of several

concepts such as domain, vertical and horizontal asymptotes, end behavior of a function, and

logarithms functions. In addition, college algebra students’ skills such as logical reasoning, use of

graph, and organization including written order, and correct use of notation and symbols were

significantly increased when they used technology. Survey of calculus students’ understanding

increased in several topics such as finding maximum/minimum for two variable functions, limits,

and definite integrals when they used technology in their class activities and on written tests.

6.2 Limitations

There are several limitations on the results of this research study. Four different teachers taught

eighth college algebra sections with different methods and knowledge and different ways to use

technology. In this research study the effect of teacher knowledge was neglected as well as the

effect of teacher methods. This research study was conducted on both female and male students

from various ethnicities. The effects of gender and ethnicity were ignored in this study.

The concept test, which was the only written test in college algebra, was designed by the

coordinator of college algebra. Therefore, the primary researcher did not have control on the design

of the tests and some of the questions were GC neutral.

The sample size in college algebra courses was large for qualitative analyses. Therefore, only

students’ understanding on some of the concepts were qualitatively analyzed.

135
This research study was conducted on survey of calculus during summer classes of five weeks,

which was the short timeframe. Students participated in class every weekday for 90 m inutes.

Therefore, students did not have enough time to go over some concepts in depth. T his could

influence students understanding of a concept. T he sample size in survey of calculus was too

small to allow a t-test to show any conclusion on RT1, RT2, SAS, and the qualitative rubric.

Both survey of calculus and college algebra courses were MLP courses, which means these courses

were more computer based, and the students’ written works were not available. The other two

review tests that were designed by the primary instructor were replacements for the students’

lowest quiz scores; therefore, students did not have strong motivation to take these tests.

Teaching based on apps requires smartphones or tablets, but some students do not have access to

this technology, which makes it difficult. Finding specific free educational apps whose designs

are based on theories of learning is not easy and requires time as well.

6.3 Implications and recommendations for future research

This research study adds to the limited experiential literature that reports on t he effect of

technology on s tudents’ understanding and organization of their written work in college level

mathematics. This study also extends information on the effect of specific technology such as

apps on students’ learning in college mathematics courses. The findings from this research study

have several implications for teaching and learning college and secondary mathematics based on

new technology. Students can be at an advantage in understanding abstract mathematical concepts

by using technologies such as online GC, handheld GC, and apps. University teachers can benefit

from new technology to enhance teaching mathematics courses. Further, teachers need access to

learning about technology-based curricula in order to teach higher mathematics courses for

conceptual understanding. Findings from this research study suggest that teachers can benefit from

136
new technology-based mathematics teaching approaches for other undergraduate mathematics

courses beyond introductory calculus.

The sample size in the survey of calculus was small and the timeframe of the research on t his

course was short. In addition, a qualitative analysis was applied to only a few topics in survey of

calculus course. Future studies could be conducted qualitatively and quantitatively on the effect of

technology on s tudents understanding of other topics of calculus with larger sample sizes and

longer timeframes.

The findings from this research study showed that college algebra students had more organized

written work and more correct use of notation and symbols when they used a GC on the test

compared to the time that they did not use a GC for similar concepts. But there is not any clear

reason why this happened, and there is not any information on the states of students’ minds. Future

research could be conducted on cognitive issues associated with the use technology and how it

affects their organization of written work.

Although in this research study data on gender and ethnicity of the participants were collected, the

effect of technology on g ender and ethnicity was not analyzed. This can be a topic for future

research and could be worthwhile information about the differences on the effect of technology on

female and male students while learning mathematics concepts.

It would take time and effort to find appropriate educational technology, tools, and apps that would

consider learning theories and specific topics from college algebra and survey of calculus.

Therefore, mathematics education research could explore new technology and tools and provide a

list of educational technologies and teaching methods that could facilitate college students learning

mathematics as well as college teachers teaching mathematics.

137
Finding from this research study showed that survey of calculus students benefited in learning of

the limit a nd integrals using GCs. In addition, survey of calculus students mentioned in their

interviews that they understand integrals and limits better when they used Desmos. Desmos has

great potential for visualizing calculus concepts such as integrals and limits and derivatives. Future

research could conduct a mixed quantitative and qualitative methods study on the effect of Desmos

on calculus students understanding that could yield valuable finding to help students to understand

calculus concepts better.

6.4 Conclusion

In this study, mixed qualitative and quantitative methods were used to investigate the effectiveness

of teaching with hand-held GCs, online GCs and smartphone applications on the understanding of

students in college algebra and survey of calculus classes. This study aimed to investigate i) what

areas of college algebra and survey of calculus are affected more by technology ii) how technology

affects the organization of students written work and, iii) the effect of technology on the attitude

of the students toward mathematics. Data were collected from different sources such as pre- and

post- student attitude surveys, scores on three written tests, and interviews with students. The

findings from this study revealed that college algebra students who used GCs had a b etter

understanding of x and y-intercept, domain of a function, end behavior, and vertical and horizontal

asymptote. However, the performance of students on function composition and word problems

was similar with or without GCs. In addition, college algebra students who used smartphone

application in their class activities were able to make a better connection between logarithms and

exponential functions. The results of the qualitative analysis showed that students’ written work is

more organized when they use technology on their tests. Students who used a GC for the test in

survey of calculus showed a better understanding of maximum and minimum for functions of two

138
variables as well as the concept of limit a nd definite integrals. However, students had similar

performance in derivative problems, indefinite integrals and limits that need the use of rules. As a

case study on the effect of technology on college level mathematics courses, this study supports

the need for future research on other undergraduate mathematics courses, especially calculus

sequences on a larger scale.

139
References

Berezovski, T., & Zazkis, R. (2006). Logarithms: Snapshots from two tasks. International Group
for the Psychology of Mathematics Education, 145.

Blair, K. P. (2013). (2013). Learning in critter corral: Evaluating three kinds of feedback in a
preschool math app. Paper presented at the Proceedings of the 12th International
Conference on Interaction Design and Children, 372-375.

Blitzer, R., Mayne, M. C., & Pietro, J. (2004). College algebra. Prentice Hall.

Bloom, B. S. (1984). The 2 sigma problem: The search for methods of group instruction as
effective as one-to-one tutoring. Educational Researcher, 13(6), 4-16.

Bookman, J., & Friedman, C. P. (1994). A comparison of the problem solving performance of
students in lab based and traditional calculus. Research in Collegiate Mathematics
Education I, 4, 101-116.

Brookstein, A., Hegedus, S., Dalton, S., Moniz, R., & Tapper, J. (2011). Measuring student
attitude in mathematics classrooms. Dartmouth: Kaput Center for Research and Innovation
in STEM Education, University of Massachusetts,

Cherner, T., Dix, J., & Lee, C. (2014). Cleaning up that mess: A framework for classifying
educational apps. Contemporary Issues in Technology and Teacher Education, 14(2), 158-
193.

Chval, K., & Khisty, L. L. (2001). (2001). Writing in mathematics with latino fifth grade
students. Paper presented at the Annual Meeting of the American Educational Research
Association, Seattle, WA,

Clark, R. E. (1994). Media will never influence learning. Educational Technology Research and
Development, 42(2), 21-29.

Connors, M. A. (1995). Achievement and gender in computer-integrated calculus. Journal of

Women and Minorities in Science and Engineering, 2(1&2)

Cooper, J. A., & Schleser, R. (2006). Closing the achievement gap: Examining the role of
cognitive developmental level in academic achievement. Early Childhood Education
Journal, 33(5), 301-306.

Creswell, J. W. (2002). Educational research: Planning, conducting, and evaluating

quantitative (pp. 146-166). Upper Saddle River, NJ: Prentice Hall.

Cunningham, R. F. (1991). The effects on achievement of using computer software to reduce

hand-generated symbolic manipulation in freshman calculus. dl.acm.org.

140
Dunham, P. H. (1995). Calculator use and gender issues. Association for Women in Mathematics
Newsletter, 25(2), 16-18.

Dunham, P. H., & Dick, T. P. (1994). Research on graphing calculators. The Mathematics
Teacher, 87(6), 440.

Ellington, A. J. (2003). A meta-analysis of the effects of calculators on students' achievement

and attitude levels in precollege mathematics classes. Journal for Research in Mathematics
Education, 433-463.

Ellington, A. J. (2006). The effects of Non‐CAS graphing calculators on student achievement

and attitude levels in mathematics: A Meta‐Analysis. School Science and Mathematics,
106(1), 16-26.

Ely, D. P. (1963). The changing role of the audiovisual process in education--a definition and a
glossary of related terms.

Falloon, G. (2013). Young students using iPads: App design and content influences on their
learning pathways. Computers & Education, 68, 505-521.

Fey, J. T., Cuoco, A., Kieran, C., McMullin, L., & Zbiek, R. M. (2003). Computer algebra
systems in secondary school mathematics education NCTM/National Council of Teachers of
Mathematics.

Glaser, R. (1976). Components of a psychology of instruction: Toward a science of design.

Review of Educational Research, 46(1), 1-24.

Goldin, G. A., & Kaput, J. J. (1996). A joint perspective on the idea of representation in learning
and doing mathematics. Theories of Mathematical Learning, , 397-430.

Green, L. S., Hechter, R. P., Tysinger, P. D., & Chassereau, K. D. (2014). Mobile app selection
for 5th through 12th grade science: The development of the MASS rubric. Computers &
Education, 75, 65-71.

Handal, B., El-Khoury, J., Campbell, C., & Cavanagh, M. (2013a). A framework for categorising
mobile applications in mathematics education.

Handal, B., El-Khoury, J., Campbell, C., & Cavanagh, M. (2013b). A framework for
categorising mobile applications in mathematics education.

Handal, B., El-Khoury, J., Campbell, C., & Cavanagh, M. (2013c). A framework for categorising
mobile applications in mathematics education.

Hargis, J., Cavanaugh, C., Kamali, T., & Soto, M. (2014). A federal higher education iPad
mobile learning initiative: Triangulation of data to determine early effectiveness. Innovative
Higher Education, 39(1), 45-57.

141
Heinich, R. (1984). The proper study of instructional technology. Educational Technology
Research and Development, 32(2), 67-88.

Heller, J. I., Curtis, D. A., Jaffe, R., & Verboncoeur, C. J. (2005). The impact of handheld
graphing calculator use on student achievement in algebra 1. Online Submission,

Henderson, S., & Yeow, J. (2012). (2012). iPad in education: A case study of iPad adoption and
use in a primary school. Paper presented at the System Science (HICSS), 2012 45th Hawaii
International Conference On, 78-87.

Highfield, K., & Goodwin, K. (2013). Apps for mathematics learning: A review
of'educational'apps from the iTunes app store.

Hoban, C. F. (1965). From theory to policy decisions. Educational Technology Research and
Development, 13(2), 121-139.

Hunter, J. S. (2011). The Effects of Graphing Calculator use on High-School Students'

Reasoning in Integral Calculus,

Januszewski, A. (2001). Educational technology: The development of a concept Libraries

Unlimited.

Jones, J. (2008). Problem-solving using graphing calculators.

Kearney, M., Schuck, S., Burden, K., & Aubusson, P. (2012). Viewing mobile learning from a
pedagogical perspective. Research in Learning Technology, 20

Kenney, R. (2005). (2005). Students’ understanding of logarithmic function notation. Paper

presented at the Annual Meeting of the North American Chapter of the International Group
for the Psychology of Mathematics Education, Hosted by Virginia Tech University Hotel
Roanoke & Conference Center, Roanoke, VA. Retrieved from Http://Www. Allacademic.
Com/Meta/p24727index. Html,

Kenney, R., & Kastberg, S. (2013). Links in learning logarithms. Australian Senior Mathematics
Journal, 27(1), 12-20.

Kinzel, M. T. (1999). Understanding algebraic notation from the students' perspective. The
Mathematics Teacher, 92(5), 436.

Kozma, R. B. (1994). Will media influence learning? reframing the debate. Educational
Technology Research and Development, 42(2), 7-19.

Krathwohl, D. R. (1956). Taxonomy of educational objectives: Cognitive domain Longmans,

Green.

142
Larkin, K. (2013). Mathematics education: Is there an app for that?. Mathematics Education
Research Group of Australasia,

Lial, M., Greenwell, R. N., & Ritchey, N. P. (2013). Calculus with applications: Pearson new
international edition Pearson Higher Ed.

Merriweather, M., & Tharp, M. L. (1999). The effect of instruction with graphing calculators on
how general mathematics students naturalistically solve algebraic problems. Journal of
Computers in Mathematics and Science Teaching, 18(1), 7-22.

Moore, M. G. (1993). 2 theory of transactional distance. Theoretical Principles of Distance

Education, , 22.

Moyer-Packenham, P. S., Shumway, J. F., Bullock, E., Tucker, S. I., Anderson-Pence, K. L.,
Westenskow, A., . . . Mahamane, S. (2015). Young children’s learning performance and
efficiency when using virtual manipulative mathematics iPad apps. Journal of Computers in
Mathematics and Science Teaching, 34(1), 41-69.

National Council of Teachers of Mathematics. (2000). Principles and standards for school
mathematics National Council of Teachers of.

National Council of Teachers of Mathematics. Commission on Standards for School

Mathematics. (1989). Curriculum and evaluation standards for school mathematics Natl
Council of Teachers of.

O'Malley, P., Jenkins, S., Wesley, B., Donehower, C., Rabuck, D., & Lewis, M. (2013).
Effectiveness of using iPads to build math fluency. Online Submission,

Park, Y. (2011). A pedagogical framework for mobile learning: Categorizing educational

applications of mobile technologies into four types. The International Review of Research in
Open and Distributed Learning, 12(2), 78-102.

Penglase, M., & Arnold, S. (1996). The graphics calculator in mathematics education: A critical
review of recent research. Mathematics Education Research Journal, 8(1), 58-90.

Piaget, J., & Inhelder, B. (1969). The psychology of the child Basic Books.

Porzio, D. T. (1997). Examining effects of graphics calculator use on students' understanding of

numerical, graphical, and symbolic representations of calculus concepts.

Quesada, A. R., & Maxwell, M. E. (1994). The effects of using graphing calculators to enhance
college students' performance in precalculus. Educational Studies in Mathematics, 27(2),
205-215.

Robinson, R., Molenda, M., & Rezabek, L. (2008). Facilitating learning. Educational
Technology: A Definition with Commentary, , 15-48.

143
Ruthven, K. (1990). The influence of graphic calculator use on translation from graphic to
symbolic forms. Educational Studies in Mathematics, 21(5), 431-450.

Sajka, M. (2003). A secondary school student's understanding of the concept of function-A case
study. Educational Studies in Mathematics, 53(3), 229-254.

Schwen, T. (1977). Professional scholarship in educational technology: Criteria for inquiry. AV

Communication Review, 25, 35-79.

Silber, K. H. (1970). What field are we in, anyhow?. Audiovisual Instr, 15(5), 21-24.

Simon, H. A. (1996). The sciences of the artificial MIT press.

Skinner, B. F. (1938). The behavior of organisms: An experimental analysis.

Skinner, B. F. (1953). Science and human behavior Simon and Schuster.

Smith, K. B., & Shotsberger, P. G. (1997). Assessing the use of graphing calculators in college
algebra: Reflecting on dimensions of teaching and learning. School Science and
Mathematics, 97(7), 368-376.

Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive
Science, 12(2), 257-285.

Sweller, J. (1994). Cognitive load theory, learning difficulty, and instructional design. Learning
and Instruction, 4(4), 295-312.

Sweller, J., & Chandler, P. (1994). Why some material is difficult to learn. Cognition and
Instruction, 12(3), 185-233.

Sweller, J., Chandler, P., Tierney, P., & Cooper, M. (1990). Cognitive load as a factor in the
structuring of technical material. Journal of Experimental Psychology: General, 119(2),
176.

Tharp, M. L., & Uprichard, A. E. (1992). A problem-solving inquiry-oriented approach to

learning mathematics: Student/teacher interactions of rule-based learners.

Traxler, J., & Kukulska-Hulme, A. (2005). Evaluating mobile learning: Reflections on current
practice.

Vygotski, L. S. (1987). The collected works of LS vygotsky.

Vygotsky, L. S. (1980). Mind in society: The development of higher psychological processes

Harvard university press.

144
Wang, Y., Wu, M., & Wang, H. (2009). Investigating the determinants and age and gender
differences in the acceptance of mobile learning. British Journal of Educational
Technology, 40(1), 92-118.

Weber, K. (2002). Students' understanding of exponential and logarithmic functions.

White-Clark, R., DiCarlo, M., & Gilchriest, S. N. (2008). " Guide on the side": An instructional
approach to meet mathematics standards. The High School Journal, 91(4), 40-44.

145
Appendix A: IRB Approval Letters

146
147
Appendix B: Pre and post-Students attitude survey Items

# Questions: SA A N/U D SD
1 Calculator should “only” be used to check work.
2 A graphing calculator can be used as a tool to solve problems I could
not solve before.
3 The graphing calculator will hinder my ability to understand basic
computation.
4 Graphing calculator make math fun.
5 Since I have a graphing calculator, I do not need to learn to make
graphs by hand.
6 Learning algebra is easier if a graphing calculator is used to solve
problems.
7 I understand mathematics better if I solve problems with pencil and
paper first before I use a graphing calculator.
8 I know how to use a graphing calculator very well.
9 It is important that everyone learn how to use a graphing calculator.
10 I would do better in math if I could use a graphing calculator.
11 I prefer working problems with a graphing calculator.
12 I am good in mathematics.
13 Mathematics is boring.
14 I would appreciate math better if I had a graphing calculator.
15 Using a graphing calculator to solve statistics problems is confusing.
16 I would try harder in math if I had a graphing calculator.
17 I feel I am cheating myself out of a chance to learn when I use a
graphing calculator.
18 If I use a graphing calculator my ability to visualize problems will
decrease.
19 Learning mathematics is mostly memorizing a set of facts and rules.
20 When doing mathematics, it is only important to know how to do a
process and not why it works.
21 Learning mathematics means exploring problems to discover patterns
and generalize.
22 I rely on my graphing calculator too much when solving problems.
23 I feel graphing calculators should not be used while taking
mathematics tests.
24- How much experience of using calculator in math courses have you had? Give your response based on a scale of

0 to 10(0 for none, 5 for 3 courses, and 10 for all of your previous math courses).

148